Datasets:

laion

/

NeurIPSX-Complete

like 0

Modelling Cellular Perturbations with the Sparse Additive Mechanism Shift Variational Autoencoder

Main Conference Track

Michael Bereket, Theofanis Karaletsos

Generative models of observations under interventions have been a vibrant topic of interest across machine learning and the sciences in recent years. For example, in drug discovery, there is a need to model the effects of diverse interventions on cells in order to characterize unknown biological mechanisms of action. We propose the Sparse Additive Mechanism Shift Variational Autoencoder, SAMS-VAE, to combine compositionality, disentanglement, and interpretability for perturbation models. SAMS-VAE models the latent state of a perturbed sample as the sum of a local latent variable capturing sample-specific variation and sparse global variables of latent intervention effects. Crucially, SAMS-VAE sparsifies these global latent variables for individual perturbations to identify disentangled, perturbation-specific latent subspaces that are flexibly composable. We evaluate SAMS-VAE both quantitatively and qualitatively on a range of tasks using two popular single cell sequencing datasets.In order to measure perturbation-specific model-properties, we also introduce a framework for evaluation of perturbation models based on average treatment effects with links to posterior predictive checks. SAMS-VAE outperforms comparable models in terms of generalization across in-distribution and out-of-distribution tasks, including a combinatorial reasoning task under resource paucity, and yields interpretable latent structures which correlate strongly to known biological mechanisms. Our results suggest SAMS-VAE is an interesting addition to the modeling toolkit for machine learning-driven scientific discovery.

https://papers.nips.cc/paper_files/paper/2023/file/0001ca33ba34ce0351e4612b744b3936-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/20165-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/0001ca33ba34ce0351e4612b744b3936-Supplemental-Conference.pdf

Cross-Episodic Curriculum for Transformer Agents

Main Conference Track

Lucy Xiaoyang Shi, Yunfan Jiang, Jake Grigsby, Linxi Fan, Yuke Zhu

We present a new algorithm, Cross-Episodic Curriculum (CEC), to boost the learning efficiency and generalization of Transformer agents. Central to CEC is the placement of cross-episodic experiences into a Transformer’s context, which forms the basis of a curriculum. By sequentially structuring online learning trials and mixed-quality demonstrations, CEC constructs curricula that encapsulate learning progression and proficiency increase across episodes. Such synergy combined with the potent pattern recognition capabilities of Transformer models delivers a powerful cross-episodic attention mechanism. The effectiveness of CEC is demonstrated under two representative scenarios: one involving multi-task reinforcement learning with discrete control, such as in DeepMind Lab, where the curriculum captures the learning progression in both individual and progressively complex settings; and the other involving imitation learning with mixed-quality data for continuous control, as seen in RoboMimic, where the curriculum captures the improvement in demonstrators' expertise. In all instances, policies resulting from CEC exhibit superior performance and strong generalization. Code is open-sourced on the project website https://cec-agent.github.io/ to facilitate research on Transformer agent learning.

https://papers.nips.cc/paper_files/paper/2023/file/001608167bb652337af5df0129aeaabd-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/22418-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/001608167bb652337af5df0129aeaabd-Supplemental-Conference.pdf

PaintSeg: Painting Pixels for Training-free Segmentation

Main Conference Track

Xiang Li, Chung-Ching Lin, Yinpeng Chen, Zicheng Liu, Jinglu Wang, Rita Singh, Bhiksha Raj

The paper introduces PaintSeg, a new unsupervised method for segmenting objects without any training. We propose an adversarial masked contrastive painting (AMCP) process, which creates a contrast between the original image and a painted image in which a masked area is painted using off-the-shelf generative models. During the painting process, inpainting and outpainting are alternated, with the former masking the foreground and filling in the background, and the latter masking the background while recovering the missing part of the foreground object. Inpainting and outpainting, also referred to as I-step and O-step, allow our method to gradually advance the target segmentation mask toward the ground truth without supervision or training. PaintSeg can be configured to work with a variety of prompts, e.g. coarse masks, boxes, scribbles, and points. Our experimental results demonstrate that PaintSeg outperforms existing approaches in coarse mask-prompt, box-prompt, and point-prompt segmentation tasks, providing a training-free solution suitable for unsupervised segmentation. Code: https://github.com/lxa9867/PaintSeg.

https://papers.nips.cc/paper_files/paper/2023/file/0021c2cb1b9b6a71ac478ea52a93b25a-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/19673-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/0021c2cb1b9b6a71ac478ea52a93b25a-Supplemental-Conference.zip

Bootstrapping Vision-Language Learning with Decoupled Language Pre-training

Main Conference Track

Yiren Jian, Chongyang Gao, Soroush Vosoughi

We present a novel methodology aimed at optimizing the application of frozen large language models (LLMs) for resource-intensive vision-language (VL) pre-training. The current paradigm uses visual features as prompts to guide language models, with a focus on determining the most relevant visual features for corresponding text. Our approach diverges by concentrating on the language component, specifically identifying the optimal prompts to align with visual features. We introduce the Prompt-Transformer (P-Former), a model that predicts these ideal prompts, which is trained exclusively on linguistic data, bypassing the need for image-text pairings. This strategy subtly bifurcates the end-to-end VL training process into an additional, separate stage. Our experiments reveal that our framework significantly enhances the performance of a robust image-to-text baseline (BLIP-2), and effectively narrows the performance gap between models trained with either 4M or 129M image-text pairs. Importantly, our framework is modality-agnostic and flexible in terms of architectural design, as validated by its successful application in a video learning task using varied base modules. The code will be made available at https://github.com/yiren-jian/BLIText.

https://papers.nips.cc/paper_files/paper/2023/file/002262941c9edfd472a79298b2ac5e17-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/20571-/bibtex

Path following algorithms for $\ell_2$-regularized $M$-estimation with approximation guarantee

Main Conference Track

Yunzhang Zhu, Renxiong Liu

Many modern machine learning algorithms are formulated as regularized M-estimation problems, in which a regularization (tuning) parameter controls a trade-off between model fit to the training data and model complexity. To select the ``best'' tuning parameter value that achieves a good trade-off, an approximated solution path needs to be computed. In practice, this is often done through selecting a grid of tuning parameter values and solving the regularized problem at the selected grid points. However, given any desired level of accuracy, it is often not clear how to choose the grid points and also how accurately one should solve the regularized problems at the selected gird points, both of which can greatly impact the overall amount of computation. In the context of $\ell_2$-regularized $M$-estimation problem, we propose a novel grid point selection scheme and an adaptive stopping criterion for any given optimization algorithm that produces an approximated solution path with approximation error guarantee. Theoretically, we prove that the proposed solution path can approximate the exact solution path to arbitrary level of accuracy, while saving the overall computation as much as possible. Numerical results also corroborate with our theoretical analysis.

https://papers.nips.cc/paper_files/paper/2023/file/00296c0e10cd24d415c2db63ea2a2c68-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/22675-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/00296c0e10cd24d415c2db63ea2a2c68-Supplemental-Conference.pdf

PDF: Point Diffusion Implicit Function for Large-scale Scene Neural Representation

Main Conference Track

Yuhan Ding, Fukun Yin, Jiayuan Fan, Hui Li, Xin Chen, Wen Liu, Chongshan Lu, Gang Yu, Tao Chen

Recent advances in implicit neural representations have achieved impressive results by sampling and fusing individual points along sampling rays in the sampling space. However, due to the explosively growing sampling space, finely representing and synthesizing detailed textures remains a challenge for unbounded large-scale outdoor scenes. To alleviate the dilemma of using individual points to perceive the entire colossal space, we explore learning the surface distribution of the scene to provide structural priors and reduce the samplable space and propose a Point Diffusion implicit Function, PDF, for large-scale scene neural representation. The core of our method is a large-scale point cloud super-resolution diffusion module that enhances the sparse point cloud reconstructed from several training images into a dense point cloud as an explicit prior. Then in the rendering stage, only sampling points with prior points within the sampling radius are retained. That is, the sampling space is reduced from the unbounded space to the scene surface. Meanwhile, to fill in the background of the scene that cannot be provided by point clouds, the region sampling based on Mip-NeRF 360 is employed to model the background representation. Expensive experiments have demonstrated the effectiveness of our method for large-scale scene novel view synthesis, which outperforms relevant state-of-the-art baselines.

https://papers.nips.cc/paper_files/paper/2023/file/0073cc73e1873b35345209b50a3dab66-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/22964-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/0073cc73e1873b35345209b50a3dab66-Supplemental-Conference.pdf

Natural Actor-Critic for Robust Reinforcement Learning with Function Approximation

Main Conference Track

Ruida Zhou, Tao Liu, Min Cheng, Dileep Kalathil, P. R. Kumar, Chao Tian

We study robust reinforcement learning (RL) with the goal of determining a well-performing policy that is robust against model mismatch between the training simulator and the testing environment. Previous policy-based robust RL algorithms mainly focus on the tabular setting under uncertainty sets that facilitate robust policy evaluation, but are no longer tractable when the number of states scales up. To this end, we propose two novel uncertainty set formulations, one based on double sampling and the other on an integral probability metric. Both make large-scale robust RL tractable even when one only has access to a simulator. We propose a robust natural actor-critic (RNAC) approach that incorporates the new uncertainty sets and employs function approximation. We provide finite-time convergence guarantees for the proposed RNAC algorithm to the optimal robust policy within the function approximation error. Finally, we demonstrate the robust performance of the policy learned by our proposed RNAC approach in multiple MuJoCo environments and a real-world TurtleBot navigation task.

https://papers.nips.cc/paper_files/paper/2023/file/007f4927e60699392425f267d43f0940-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/19552-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/007f4927e60699392425f267d43f0940-Supplemental-Conference.zip

Adaptive Selective Sampling for Online Prediction with Experts

Main Conference Track

Rui Castro, Fredrik Hellström, Tim van Erven

We consider online prediction of a binary sequence with expert advice. For this setting, we devise label-efficient forecasting algorithms, which use a selective sampling scheme that enables collecting much fewer labels than standard procedures. For the general case without a perfect expert, we prove best-of-both-worlds guarantees, demonstrating that the proposed forecasting algorithm always queries sufficiently many labels in the worst case to obtain optimal regret guarantees, while simultaneously querying much fewer labels in more benign settings. Specifically, for a scenario where one expert is strictly better than the others in expectation, we show that the label complexity of the label-efficient forecaster is roughly upper-bounded by the square root of the number of rounds. Finally, we present numerical experiments empirically showing that the normalized regret of the label-efficient forecaster can asymptotically match known minimax rates for pool-based active learning, suggesting it can optimally adapt to benign settings.

https://papers.nips.cc/paper_files/paper/2023/file/00b67df24009747e8bbed4c2c6f9c825-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/19791-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/00b67df24009747e8bbed4c2c6f9c825-Supplemental-Conference.zip

Gigastep - One Billion Steps per Second Multi-agent Reinforcement Learning

Datasets and Benchmarks Track

Mathias Lechner, lianhao yin, Tim Seyde, Tsun-Hsuan Johnson Wang, Wei Xiao, Ramin Hasani, Joshua Rountree, Daniela Rus

Multi-agent reinforcement learning (MARL) research is faced with a trade-off: it either uses complex environments requiring large compute resources, which makes it inaccessible to researchers with limited resources, or relies on simpler dynamics for faster execution, which makes the transferability of the results to more realistic tasks challenging. Motivated by these challenges, we present Gigastep, a fully vectorizable, MARL environment implemented in JAX, capable of executing up to one billion environment steps per second on consumer-grade hardware. Its design allows for comprehensive MARL experimentation, including a complex, high-dimensional space defined by 3D dynamics, stochasticity, and partial observations. Gigastep supports both collaborative and adversarial tasks, continuous and discrete action spaces, and provides RGB image and feature vector observations, allowing the evaluation of a wide range of MARL algorithms. We validate Gigastep's usability through an extensive set of experiments, underscoring its role in widening participation and promoting inclusivity in the MARL research community.

https://papers.nips.cc/paper_files/paper/2023/file/00ba06ba5c324efdfb068865ca44cf0b-Paper-Datasets_and_Benchmarks.pdf

https://papers.nips.cc/paper_files/paper/22366-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/00ba06ba5c324efdfb068865ca44cf0b-Supplemental-Datasets_and_Benchmarks.pdf

Attentive Transfer Entropy to Exploit Transient Emergence of Coupling Effect

Main Conference Track

Xiaolei Ru, XINYA ZHANG, Zijia Liu, Jack Murdoch Moore, Gang Yan

We consider the problem of reconstructing coupled networks (e.g., biological neural networks) connecting large numbers of variables (e.g.,nerve cells), of which state evolution is governed by dissipative dynamics consisting of strong self-drive (dominants the evolution) and weak coupling-drive. The core difficulty is sparseness of coupling effect that emerges (the coupling force is significant) only momentarily and otherwise remains quiescent in time series (e.g., neuronal activity sequence). Here we learn the idea from attention mechanism to guide the classifier to make inference focusing on the critical regions of time series data where coupling effect may manifest. Specifically, attention coefficients are assigned autonomously by artificial neural networks trained to maximise the Attentive Transfer Entropy (ATEn), which is a novel generalization of the iconic transfer entropy metric. Our results show that, without any prior knowledge of dynamics, ATEn explicitly identifies areas where the strength of coupling-drive is distinctly greater than zero. This innovation substantially improves reconstruction performance for both synthetic and real directed coupling networks using data generated by neuronal models widely used in neuroscience.

https://papers.nips.cc/paper_files/paper/2023/file/00bb4e415ef117f2dee2fc3b778d806d-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/21975-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/00bb4e415ef117f2dee2fc3b778d806d-Supplemental-Conference.zip

PopSign ASL v1.0: An Isolated American Sign Language Dataset Collected via Smartphones

Datasets and Benchmarks Track

Thad Starner, Sean Forbes, Matthew So, David Martin, Rohit Sridhar, Gururaj Deshpande, Sam Sepah, Sahir Shahryar, Khushi Bhardwaj, Tyler Kwok, Daksh Sehgal, Saad Hassan, Bill Neubauer, Sofia Vempala, Alec Tan, Jocelyn Heath, Unnathi Kumar, Priyanka Mosur, Tavenner Hall, Rajandeep Singh, Christopher Cui, Glenn Cameron, Sohier Dane, Garrett Tanzer

PopSign is a smartphone-based bubble-shooter game that helps hearing parentsof deaf infants learn sign language. To help parents practice their ability to sign,PopSign is integrating sign language recognition as part of its gameplay. Fortraining the recognizer, we introduce the PopSign ASL v1.0 dataset that collectsexamples of 250 isolated American Sign Language (ASL) signs using Pixel 4Asmartphone selfie cameras in a variety of environments. It is the largest publiclyavailable, isolated sign dataset by number of examples and is the first dataset tofocus on one-handed, smartphone signs. We collected over 210,000 examplesat 1944x2592 resolution made by 47 consenting Deaf adult signers for whomAmerican Sign Language is their primary language. We manually reviewed 217,866of these examples, of which 175,023 (approximately 700 per sign) were the signintended for the educational game. 39,304 examples were recognizable as a signbut were not the desired variant or were a different sign. We provide a training setof 31 signers, a validation set of eight signers, and a test set of eight signers. Abaseline LSTM model for the 250-sign vocabulary achieves 82.1% accuracy (81.9%class-weighted F1 score) on the validation set and 84.2% (83.9% class-weightedF1 score) on the test set. Gameplay suggests that accuracy will be sufficient forcreating educational games involving sign language recognition.

https://papers.nips.cc/paper_files/paper/2023/file/00dada608b8db212ea7d9d92b24c68de-Paper-Datasets_and_Benchmarks.pdf

https://papers.nips.cc/paper_files/paper/22632-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/00dada608b8db212ea7d9d92b24c68de-Supplemental-Datasets_and_Benchmarks.pdf

Provable Adversarial Robustness for Group Equivariant Tasks: Graphs, Point Clouds, Molecules, and More

Main Conference Track

Jan Schuchardt, Yan Scholten, Stephan Günnemann

A machine learning model is traditionally considered robust if its prediction remains (almost) constant under input perturbations with small norm. However, real-world tasks like molecular property prediction or point cloud segmentation have inherent equivariances, such as rotation or permutation equivariance. In such tasks, even perturbations with large norm do not necessarily change an input's semantic content. Furthermore, there are perturbations for which a model's prediction explicitly needs to change. For the first time, we propose a sound notion of adversarial robustness that accounts for task equivariance. We then demonstrate that provable robustness can be achieved by (1) choosing a model that matches the task's equivariances (2) certifying traditional adversarial robustness. Certification methods are, however, unavailable for many models, such as those with continuous equivariances. We close this gap by developing the framework of equivariance-preserving randomized smoothing, which enables architecture-agnostic certification. We additionally derive the first architecture-specific graph edit distance certificates, i.e. sound robustness guarantees for isomorphism equivariant tasks like node classification. Overall, a sound notion of robustness is an important prerequisite for future work at the intersection of robust and geometric machine learning.

https://papers.nips.cc/paper_files/paper/2023/file/00db17c36b5435195760520efa96d99c-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/21081-/bibtex

Self-Supervised Motion Magnification by Backpropagating Through Optical Flow

Main Conference Track

Zhaoying Pan, Daniel Geng, Andrew Owens

This paper presents a simple, self-supervised method for magnifying subtle motions in video: given an input video and a magnification factor, we manipulate the video such that its new optical flow is scaled by the desired amount. To train our model, we propose a loss function that estimates the optical flow of the generated video and penalizes how far if deviates from the given magnification factor. Thus, training involves differentiating through a pretrained optical flow network. Since our model is self-supervised, we can further improve its performance through test-time adaptation, by finetuning it on the input video. It can also be easily extended to magnify the motions of only user-selected objects. Our approach avoids the need for synthetic magnification datasets that have been used to train prior learning-based approaches. Instead, it leverages the existing capabilities of off-the-shelf motion estimators. We demonstrate the effectiveness of our method through evaluations of both visual quality and quantitative metrics on a range of real-world and synthetic videos, and we show our method works for both supervised and unsupervised optical flow methods.

https://papers.nips.cc/paper_files/paper/2023/file/00ed9ab006311be67879ecef8f80d7c5-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/22090-/bibtex

TexQ: Zero-shot Network Quantization with Texture Feature Distribution Calibration

Main Conference Track

Xinrui Chen, Yizhi Wang, Renao YAN, Yiqing Liu, Tian Guan, Yonghong He

Quantization is an effective way to compress neural networks. By reducing the bit width of the parameters, the processing efficiency of neural network models at edge devices can be notably improved. Most conventional quantization methods utilize real datasets to optimize quantization parameters and fine-tune. Due to the inevitable privacy and security issues of real samples, the existing real-data-driven methods are no longer applicable. Thus, a natural method is to introduce synthetic samples for zero-shot quantization (ZSQ). However, the conventional synthetic samples fail to retain the detailed texture feature distributions, which severely limits the knowledge transfer and performance of the quantized model. In this paper, a novel ZSQ method, TexQ is proposed to address this issue. We first synthesize a calibration image and extract its calibration center for each class with a texture feature energy distribution calibration method. Then, the calibration centers are used to guide the generator to synthesize samples. Finally, we introduce the mixup knowledge distillation module to diversify synthetic samples for fine-tuning. Extensive experiments on CIFAR10/100 and ImageNet show that TexQ is observed to perform state-of-the-art in ultra-low bit width quantization. For example, when ResNet-18 is quantized to 3-bit, TexQ achieves a 12.18% top-1 accuracy increase on ImageNet compared to state-of-the-art methods. Code at https://github.com/dangsingrue/TexQ.

https://papers.nips.cc/paper_files/paper/2023/file/0113ef4642264adc2e6924a3cbbdf532-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/20087-/bibtex

Ambient Diffusion: Learning Clean Distributions from Corrupted Data

Main Conference Track

Giannis Daras, Kulin Shah, Yuval Dagan, Aravind Gollakota, Alex Dimakis, Adam Klivans

We present the first diffusion-based framework that can learn an unknown distribution using only highly-corrupted samples. This problem arises in scientific applications where access to uncorrupted samples is impossible or expensive to acquire. Another benefit of our approach is the ability to train generative models that are less likely to memorize any individual training sample, since they never observe clean training data. Our main idea is to introduce additional measurement distortion during the diffusion process and require the model to predict the original corrupted image from the further corrupted image. We prove that our method leads to models that learn the conditional expectation of the full uncorrupted image given this additional measurement corruption. This holds for any corruption process that satisfies some technical conditions (and in particular includes inpainting and compressed sensing). We train models on standard benchmarks (CelebA, CIFAR-10 and AFHQ) and show that we can learn the distribution even when all the training samples have 90\% of their pixels missing. We also show that we can finetune foundation models on small corrupted datasets (e.g. MRI scans with block corruptions) and learn the clean distribution without memorizing the training set.

https://papers.nips.cc/paper_files/paper/2023/file/012af729c5d14d279581fc8a5db975a1-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/21484-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/012af729c5d14d279581fc8a5db975a1-Supplemental-Conference.zip

Scalable Membership Inference Attacks via Quantile Regression

Main Conference Track

Martin Bertran, Shuai Tang, Aaron Roth, Michael Kearns, Jamie H. Morgenstern, Steven Z. Wu

Membership inference attacks are designed to determine, using black box access to trained models, whether a particular example was used in training or not. Membership inference can be formalized as a hypothesis testing problem. The most effective existing attacks estimate the distribution of some test statistic (usually the model's confidence on the true label) on points that were (and were not) used in training by training many \emph{shadow models}---i.e. models of the same architecture as the model being attacked, trained on a random subsample of data. While effective, these attacks are extremely computationally expensive, especially when the model under attack is large. \footnotetext[0]{Martin and Shuai are the lead authors, and other authors are ordered alphabetically. {maberlop,shuat}@amazon.com}We introduce a new class of attacks based on performing quantile regression on the distribution of confidence scores induced by the model under attack on points that are not used in training. We show that our method is competitive with state-of-the-art shadow model attacks, while requiring substantially less compute because our attack requires training only a single model. Moreover, unlike shadow model attacks, our proposed attack does not require any knowledge of the architecture of the model under attack and is therefore truly ``black-box". We show the efficacy of this approach in an extensive series of experiments on various datasets and model architectures. Our code is available at \href{https://github.com/amazon-science/quantile-mia}{github.com/amazon-science/quantile-mia.}

https://papers.nips.cc/paper_files/paper/2023/file/01328d0767830e73a612f9073e9ff15f-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/20306-/bibtex

ESSEN: Improving Evolution State Estimation for Temporal Networks using Von Neumann Entropy

Main Conference Track

Qiyao Huang, Yingyue Zhang, Zhihong Zhang, Edwin Hancock

Temporal networks are widely used as abstract graph representations for real-world dynamic systems. Indeed, recognizing the network evolution states is crucial in understanding and analyzing temporal networks. For instance, social networks will generate the clustering and formation of tightly-knit groups or communities over time, relying on the triadic closure theory. However, the existing methods often struggle to account for the time-varying nature of these network structures, hindering their performance when applied to networks with complex evolution states. To mitigate this problem, we propose a novel framework called ESSEN, an Evolution StateS awarE Network, to measure temporal network evolution using von Neumann entropy and thermodynamic temperature. The developed framework utilizes a von Neumann entropy aware attention mechanism and network evolution state contrastive learning in the graph encoding. In addition, it employs a unique decoder the so-called Mixture of Thermodynamic Experts (MoTE) for decoding. ESSEN extracts local and global network evolution information using thermodynamic features and adaptively recognizes the network evolution states. Moreover, the proposed method is evaluated on link prediction tasks under both transductive and inductive settings, with the corresponding results demonstrating its effectiveness compared to various state-of-the-art baselines.

https://papers.nips.cc/paper_files/paper/2023/file/0147d967a5db3b8dde08d2a327b24568-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/19868-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/0147d967a5db3b8dde08d2a327b24568-Supplemental-Conference.pdf

Label Correction of Crowdsourced Noisy Annotations with an Instance-Dependent Noise Transition Model

Main Conference Track

Hui GUO, Boyu Wang, Grace Yi

The predictive ability of supervised learning algorithms hinges on the quality of annotated examples, whose labels often come from multiple crowdsourced annotators with diverse expertise. To aggregate noisy crowdsourced annotations, many existing methods employ an annotator-specific instance-independent noise transition matrix to characterize the labeling skills of each annotator. Learning an instance-dependent noise transition model, however, is challenging and remains relatively less explored. To address this problem, in this paper, we formulate the noise transition model in a Bayesian framework and subsequently design a new label correction algorithm. Specifically, we approximate the instance-dependent noise transition matrices using a Bayesian network with a hierarchical spike and slab prior. To theoretically characterize the distance between the noise transition model and the true instance-dependent noise transition matrix, we provide a posterior-concentration theorem that ensures the posterior consistency in terms of the Hellinger distance. We further formulate the label correction process as a hypothesis testing problem and propose a novel algorithm to infer the true label from the noisy annotations based on the pairwise likelihood ratio test. Moreover, we establish an information-theoretic bound on the Bayes error for the proposed method. We validate the effectiveness of our approach through experiments on benchmark and real-world datasets.

https://papers.nips.cc/paper_files/paper/2023/file/015a8c69bedcb0a7b2ed2e1678f34399-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/20354-/bibtex

Diffused Task-Agnostic Milestone Planner

Main Conference Track

Mineui Hong, Minjae Kang, Songhwai Oh

Addressing decision-making problems using sequence modeling to predict future trajectories shows promising results in recent years.In this paper, we take a step further to leverage the sequence predictive method in wider areas such as long-term planning, vision-based control, and multi-task decision-making.To this end, we propose a method to utilize a diffusion-based generative sequence model to plan a series of milestones in a latent space and to have an agent to follow the milestones to accomplish a given task.The proposed method can learn control-relevant, low-dimensional latent representations of milestones, which makes it possible to efficiently perform long-term planning and vision-based control.Furthermore, our approach exploits generation flexibility of the diffusion model, which makes it possible to plan diverse trajectories for multi-task decision-making.We demonstrate the proposed method across offline reinforcement learning (RL) benchmarks and an visual manipulation environment.The results show that our approach outperforms offline RL methods in solving long-horizon, sparse-reward tasks and multi-task problems,while also achieving the state-of-the-art performance on the most challenging vision-based manipulation benchmark.

https://papers.nips.cc/paper_files/paper/2023/file/0163ca1c69f848e766cfb0b7bb7e17f4-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/20632-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/0163ca1c69f848e766cfb0b7bb7e17f4-Supplemental-Conference.zip

Task-aware Distributed Source Coding under Dynamic Bandwidth

Main Conference Track

Po-han Li, Sravan Kumar Ankireddy, Ruihan (Philip) Zhao, Hossein Nourkhiz Mahjoub, Ehsan Moradi Pari, Ufuk Topcu, Sandeep Chinchali, Hyeji Kim

Efficient compression of correlated data is essential to minimize communication overload in multi-sensor networks. In such networks, each sensor independently compresses the data and transmits them to a central node. A decoder at the central node decompresses and passes the data to a pre-trained machine learning-based task model to generate the final output. Due to limited communication bandwidth, it is important for the compressor to learn only the features that are relevant to the task. Additionally, the final performance depends heavily on the total available bandwidth. In practice, it is common to encounter varying availability in bandwidth. Since higher bandwidth results in better performance, it is essential for the compressor to dynamically take advantage of the maximum available bandwidth at any instant. In this work, we propose a novel distributed compression framework composed of independent encoders and a joint decoder, which we call neural distributed principal component analysis (NDPCA). NDPCA flexibly compresses data from multiple sources to any available bandwidth with a single model, reducing compute and storage overhead. NDPCA achieves this by learning low-rank task representations and efficiently distributing bandwidth among sensors, thus providing a graceful trade-off between performance and bandwidth. Experiments show that NDPCA improves the success rate of multi-view robotic arm manipulation by 9% and the accuracy of object detection tasks on satellite imagery by 14% compared to an autoencoder with uniform bandwidth allocation.

https://papers.nips.cc/paper_files/paper/2023/file/016c63403370d81c24c1ca0123de6cfa-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/20137-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/016c63403370d81c24c1ca0123de6cfa-Supplemental-Conference.pdf

BubbleML: A Multiphase Multiphysics Dataset and Benchmarks for Machine Learning

Datasets and Benchmarks Track

Sheikh Md Shakeel Hassan, Arthur Feeney, Akash Dhruv, Jihoon Kim, Youngjoon Suh, Jaiyoung Ryu, Yoonjin Won, Aparna Chandramowlishwaran

In the field of phase change phenomena, the lack of accessible and diverse datasets suitable for machine learning (ML) training poses a significant challenge. Existing experimental datasets are often restricted, with limited availability and sparse ground truth, impeding our understanding of this complex multiphysics phenomena. To bridge this gap, we present the BubbleML dataset which leverages physics-driven simulations to provide accurate ground truth information for various boiling scenarios, encompassing nucleate pool boiling, flow boiling, and sub-cooled boiling. This extensive dataset covers a wide range of parameters, including varying gravity conditions, flow rates, sub-cooling levels, and wall superheat, comprising 79 simulations. BubbleML is validated against experimental observations and trends, establishing it as an invaluable resource for ML research. Furthermore, we showcase its potential to facilitate the exploration of diverse downstream tasks by introducing two benchmarks: (a) optical flow analysis to capture bubble dynamics, and (b) neural PDE solvers for learning temperature and flow dynamics. The BubbleML dataset and its benchmarks aim to catalyze progress in ML-driven research on multiphysics phase change phenomena, providing robust baselines for the development and comparison of state-of-the-art techniques and models.

https://papers.nips.cc/paper_files/paper/2023/file/01726ae05d72ddba3ac784a5944fa1ef-Paper-Datasets_and_Benchmarks.pdf

https://papers.nips.cc/paper_files/paper/20423-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/01726ae05d72ddba3ac784a5944fa1ef-Supplemental-Datasets_and_Benchmarks.pdf

ANTN: Bridging Autoregressive Neural Networks and Tensor Networks for Quantum Many-Body Simulation

Main Conference Track

Zhuo Chen, Laker Newhouse, Eddie Chen, Di Luo, Marin Soljacic

Quantum many-body physics simulation has important impacts on understanding fundamental science and has applications to quantum materials design and quantum technology. However, due to the exponentially growing size of the Hilbert space with respect to the particle number, a direct simulation is intractable. While representing quantum states with tensor networks and neural networks are the two state-of-the-art methods for approximate simulations, each has its own limitations in terms of expressivity and inductive bias. To address these challenges, we develop a novel architecture, Autoregressive Neural TensorNet (ANTN), which bridges tensor networks and autoregressive neural networks. We show that Autoregressive Neural TensorNet parameterizes normalized wavefunctions, allows for exact sampling, generalizes the expressivity of tensor networks and autoregressive neural networks, and inherits a variety of symmetries from autoregressive neural networks. We demonstrate our approach on quantum state learning as well as finding the ground state of the challenging 2D $J_1$-$J_2$ Heisenberg model with different systems sizes and coupling parameters, outperforming both tensor networks and autoregressive neural networks. Our work opens up new opportunities for quantum many-body physics simulation, quantum technology design, and generative modeling in artificial intelligence.

https://papers.nips.cc/paper_files/paper/2023/file/01772a8b0420baec00c4d59fe2fbace6-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/19514-/bibtex

Causal Effect Identification in Uncertain Causal Networks

Main Conference Track

Sina Akbari, Fateme Jamshidi, Ehsan Mokhtarian, Matthew Vowels, Jalal Etesami, Negar Kiyavash

Causal identification is at the core of the causal inference literature, where complete algorithms have been proposed to identify causal queries of interest. The validity of these algorithms hinges on the restrictive assumption of having access to a correctly specified causal structure. In this work, we study the setting where a probabilistic model of the causal structure is available. Specifically, the edges in a causal graph exist with uncertainties which may, for example, represent degree of belief from domain experts. Alternatively, the uncertainty about an edge may reflect the confidence of a particular statistical test. The question that naturally arises in this setting is: Given such a probabilistic graph and a specific causal effect of interest, what is the subgraph which has the highest plausibility and for which the causal effect is identifiable? We show that answering this question reduces to solving an NP-hard combinatorial optimization problem which we call the edge ID problem. We propose efficient algorithms to approximate this problem and evaluate them against both real-world networks and randomly generated graphs.

https://papers.nips.cc/paper_files/paper/2023/file/017c897b4d85a744f345ccbf9d71e501-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/21831-/bibtex

FAST: a Fused and Accurate Shrinkage Tree for Heterogeneous Treatment Effects Estimation

Main Conference Track

Jia Gu, Caizhi Tang, Han Yan, Qing Cui, Longfei Li, Jun Zhou

This paper proposes a novel strategy for estimating the heterogeneous treatment effect called the Fused and Accurate Shrinkage Tree ($\mathrm{FAST}$). Our approach utilizes both trial and observational data to improve the accuracy and robustness of the estimator. Inspired by the concept of shrinkage estimation in statistics, we develop an optimal weighting scheme and a corresponding estimator that balances the unbiased estimator based on the trial data with the potentially biased estimator based on the observational data. Specifically, combined with tree-based techniques, we introduce a new split criterion that utilizes both trial data and observational data to more accurately estimate the treatment effect. Furthermore, we confirm the consistency of our proposed tree-based estimator and demonstrate the effectiveness of our criterion in reducing prediction error through theoretical analysis. The advantageous finite sample performance of the $\mathrm{FAST}$ and its ensemble version over existing methods is demonstrated via simulations and real data analysis.

https://papers.nips.cc/paper_files/paper/2023/file/01830c92c6558179fa6d7fb1edff692c-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/20615-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/01830c92c6558179fa6d7fb1edff692c-Supplemental-Conference.pdf

Characterizing Graph Datasets for Node Classification: Homophily-Heterophily Dichotomy and Beyond

Main Conference Track

Oleg Platonov, Denis Kuznedelev, Artem Babenko, Liudmila Prokhorenkova

Homophily is a graph property describing the tendency of edges to connect similar nodes; the opposite is called heterophily. It is often believed that heterophilous graphs are challenging for standard message-passing graph neural networks (GNNs), and much effort has been put into developing efficient methods for this setting. However, there is no universally agreed-upon measure of homophily in the literature. In this work, we show that commonly used homophily measures have critical drawbacks preventing the comparison of homophily levels across different datasets. For this, we formalize desirable properties for a proper homophily measure and verify which measures satisfy which properties. In particular, we show that a measure that we call adjusted homophily satisfies more desirable properties than other popular homophily measures while being rarely used in graph machine learning literature. Then, we go beyond the homophily-heterophily dichotomy and propose a new characteristic that allows one to further distinguish different sorts of heterophily. The proposed label informativeness (LI) characterizes how much information a neighbor's label provides about a node's label. We prove that this measure satisfies important desirable properties. We also observe empirically that LI better agrees with GNN performance compared to homophily measures, which confirms that it is a useful characteristic of the graph structure.

https://papers.nips.cc/paper_files/paper/2023/file/01b681025fdbda8e935a66cc5bb6e9de-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/20330-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/01b681025fdbda8e935a66cc5bb6e9de-Supplemental-Conference.zip

Equivariant Flow Matching with Hybrid Probability Transport for 3D Molecule Generation

Main Conference Track

Yuxuan Song, Jingjing Gong, Minkai Xu, Ziyao Cao, Yanyan Lan, Stefano Ermon, Hao Zhou, Wei-Ying Ma

The generation of 3D molecules requires simultaneously deciding the categorical features (atom types) and continuous features (atom coordinates). Deep generative models, especially Diffusion Models (DMs), have demonstrated effectiveness in generating feature-rich geometries. However, existing DMs typically suffer from unstable probability dynamics with inefficient sampling speed. In this paper, we introduce geometric flow matching, which enjoys the advantages of both equivariant modeling and stabilized probability dynamics. More specifically, we propose a hybrid probability path where the coordinates probability path is regularized by an equivariant optimal transport, and the information between different modalities is aligned. Experimentally, the proposed method could consistently achieve better performance on multiple molecule generation benchmarks with 4.75$\times$ speed up of sampling on average.

https://papers.nips.cc/paper_files/paper/2023/file/01d64478381c33e29ed611f1719f5a37-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/22270-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/01d64478381c33e29ed611f1719f5a37-Supplemental-Conference.zip

Hyperbolic VAE via Latent Gaussian Distributions

Main Conference Track

Seunghyuk Cho, Juyong Lee, Dongwoo Kim

We propose a Gaussian manifold variational auto-encoder (GM-VAE) whose latent space consists of a set of Gaussian distributions. It is known that the set of the univariate Gaussian distributions with the Fisher information metric form a hyperbolic space, which we call a Gaussian manifold. To learn the VAE endowed with the Gaussian manifolds, we propose a pseudo-Gaussian manifold normal distribution based on the Kullback-Leibler divergence, a local approximation of the squared Fisher-Rao distance, to define a density over the latent space. We demonstrate the efficacy of GM-VAE on two different tasks: density estimation of image datasets and state representation learning for model-based reinforcement learning. GM-VAE outperforms the other variants of hyperbolic- and Euclidean-VAEs on density estimation tasks and shows competitive performance in model-based reinforcement learning. We observe that our model provides strong numerical stability, addressing a common limitation reported in previous hyperbolic-VAEs. The implementation is available at https://github.com/ml-postech/GM-VAE.

https://papers.nips.cc/paper_files/paper/2023/file/01ecd39ca49ddecc5729ca996304781b-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/22775-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/01ecd39ca49ddecc5729ca996304781b-Supplemental-Conference.zip

A Simple Solution for Offline Imitation from Observations and Examples with Possibly Incomplete Trajectories

Main Conference Track

Kai Yan, Alex Schwing, Yu-Xiong Wang

Offline imitation from observations aims to solve MDPs where only task-specific expert states and task-agnostic non-expert state-action pairs are available. Offline imitation is useful in real-world scenarios where arbitrary interactions are costly and expert actions are unavailable. The state-of-the-art ‘DIstribution Correction Estimation’ (DICE) methods minimize divergence of state occupancy between expert and learner policies and retrieve a policy with weighted behavior cloning; however, their results are unstable when learning from incomplete trajectories, due to a non-robust optimization in the dual domain. To address the issue, in this paper, we propose Trajectory-Aware Imitation Learning from Observations (TAILO). TAILO uses a discounted sum along the future trajectory as the weight for weighted behavior cloning. The terms for the sum are scaled by the output of a discriminator, which aims to identify expert states. Despite simplicity, TAILO works well if there exist trajectories or segments of expert behavior in the task-agnostic data, a common assumption in prior work. In experiments across multiple testbeds, we find TAILO to be more robust and effective, particularly with incomplete trajectories.

https://papers.nips.cc/paper_files/paper/2023/file/0203f489345567b4a048c38f507cdbfa-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/20292-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/0203f489345567b4a048c38f507cdbfa-Supplemental-Conference.zip

Defending against Data-Free Model Extraction by Distributionally Robust Defensive Training

Main Conference Track

Zhenyi Wang, Li Shen, Tongliang Liu, Tiehang Duan, Yanjun Zhu, Donglin Zhan, DAVID DOERMANN, Mingchen Gao

Data-Free Model Extraction (DFME) aims to clone a black-box model without knowing its original training data distribution, making it much easier for attackers to steal commercial models. Defense against DFME faces several challenges: (i) effectiveness; (ii) efficiency; (iii) no prior on the attacker's query data distribution and strategy. However, existing defense methods: (1) are highly computation and memory inefficient; or (2) need strong assumptions about attack data distribution; or (3) can only delay the attack or prove a model theft after the model stealing has happened. In this work, we propose a Memory and Computation efficient defense approach, named MeCo, to prevent DFME from happening while maintaining the model utility simultaneously by distributionally robust defensive training on the target victim model. Specifically, we randomize the input so that it: (1) causes a mismatch of the knowledge distillation loss for attackers; (2) disturbs the zeroth-order gradient estimation; (3) changes the label prediction for the attack query data. Therefore, the attacker can only extract misleading information from the black-box model. Extensive experiments on defending against both decision-based and score-based DFME demonstrate that MeCo can significantly reduce the effectiveness of existing DFME methods and substantially improve running efficiency.

https://papers.nips.cc/paper_files/paper/2023/file/0207c9ea9faf66c6e892c3fa3c167b75-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/22026-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/0207c9ea9faf66c6e892c3fa3c167b75-Supplemental-Conference.zip

Large language models transition from integrating across position-yoked, exponential windows to structure-yoked, power-law windows

Main Conference Track

David Skrill, Samuel Norman-Haignere

Modern language models excel at integrating across long temporal scales needed to encode linguistic meaning and show non-trivial similarities to biological neural systems. Prior work suggests that human brain responses to language exhibit hierarchically organized "integration windows" that substantially constrain the overall influence of an input token (e.g., a word) on the neural response. However, little prior work has attempted to use integration windows to characterize computations in large language models (LLMs). We developed a simple word-swap procedure for estimating integration windows from black-box language models that does not depend on access to gradients or knowledge of the model architecture (e.g., attention weights). Using this method, we show that trained LLMs exhibit stereotyped integration windows that are well-fit by a convex combination of an exponential and a power-law function, with a partial transition from exponential to power-law dynamics across network layers. We then introduce a metric for quantifying the extent to which these integration windows vary with structural boundaries (e.g., sentence boundaries), and using this metric, we show that integration windows become increasingly yoked to structure at later network layers. None of these findings were observed in an untrained model, which as expected integrated uniformly across its input. These results suggest that LLMs learn to integrate information in natural language using a stereotyped pattern: integrating across position-yoked, exponential windows at early layers, followed by structure-yoked, power-law windows at later layers. The methods we describe in this paper provide a general-purpose toolkit for understanding temporal integration in language models, facilitating cross-disciplinary research at the intersection of biological and artificial intelligence.

https://papers.nips.cc/paper_files/paper/2023/file/020ad0ac6a1974e6748e4a5a48110a07-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/22176-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/020ad0ac6a1974e6748e4a5a48110a07-Supplemental-Conference.pdf

Where are we in the search for an Artificial Visual Cortex for Embodied Intelligence?

Main Conference Track

Arjun Majumdar, Karmesh Yadav, Sergio Arnaud, Jason Ma, Claire Chen, Sneha Silwal, Aryan Jain, Vincent-Pierre Berges, Tingfan Wu, Jay Vakil, Pieter Abbeel, Jitendra Malik, Dhruv Batra, Yixin Lin, Oleksandr Maksymets, Aravind Rajeswaran, Franziska Meier

We present the largest and most comprehensive empirical study of pre-trained visual representations (PVRs) or visual ‘foundation models’ for Embodied AI. First, we curate CortexBench, consisting of 17 different tasks spanning locomotion, navigation, dexterous, and mobile manipulation. Next, we systematically evaluate existing PVRs and find that none are universally dominant. To study the effect of pre-training data size and diversity, we combine over 4,000 hours of egocentric videos from 7 different sources (over 4.3M images) and ImageNet to train different-sized vision transformers using Masked Auto-Encoding (MAE) on slices of this data. Contrary to inferences from prior work, we find that scaling dataset size and diversity does not improve performance universally (but does so on average). Our largest model, named VC-1, outperforms all prior PVRs on average but does not universally dominate either. Next, we show that task- or domain-specific adaptation of VC-1 leads to substantial gains, with VC-1 (adapted) achieving competitive or superior performance than the best known results on all of the benchmarks in CortexBench. Finally, we present real-world hardware experiments, in which VC-1 and VC-1 (adapted) outperform the strongest pre-existing PVR. Overall, this paper presents no new techniques but a rigorous systematic evaluation, a broad set of findings about PVRs (that in some cases, refute those made in narrow domains in prior work), and open-sourced code and models (that required over 10,000 GPU-hours to train) for the benefit of the research community.

https://papers.nips.cc/paper_files/paper/2023/file/022ca1bed6b574b962c48a2856eb207b-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/20933-/bibtex

Belief Projection-Based Reinforcement Learning for Environments with Delayed Feedback

Main Conference Track

Jangwon Kim, Hangyeol Kim, Jiwook Kang, Jongchan Baek, Soohee Han

We present a novel actor-critic algorithm for an environment with delayed feedback, which addresses the state-space explosion problem of conventional approaches. Conventional approaches use an augmented state constructed from the last observed state and actions executed since visiting the last observed state. Using the augmented state space, the correct Markov decision process for delayed environments can be constructed; however, this causes the state space to explode as the number of delayed timesteps increases, leading to slow convergence. Our proposed algorithm, called Belief-Projection-Based Q-learning (BPQL), addresses the state-space explosion problem by evaluating the values of the critic for which the input state size is equal to the original state-space size rather than that of the augmented one. We compare BPQL to traditional approaches in continuous control tasks and demonstrate that it significantly outperforms other algorithms in terms of asymptotic performance and sample efficiency. We also show that BPQL solves long-delayed environments, which conventional approaches are unable to do.

https://papers.nips.cc/paper_files/paper/2023/file/0252a434b18962c94910c07cd9a7fecc-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/21787-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/0252a434b18962c94910c07cd9a7fecc-Supplemental-Conference.zip

Batchnorm Allows Unsupervised Radial Attacks

Main Conference Track

Amur Ghose, Apurv Gupta, Yaoliang Yu, Pascal Poupart

The construction of adversarial examples usually requires the existence of soft or hard labels for each instance, with respect to which a loss gradient provides the signal for construction of the example. We show that for batch normalized deep image recognition architectures, intermediate latents that are produced after a batch normalization step by themselves suffice to produce adversarial examples using an intermediate loss solely utilizing angular deviations, without relying on any label. We motivate our loss through the geometry of batch normed representations and their concentration of norm on a hypersphere and distributional proximity to Gaussians. Our losses expand intermediate latent based attacks that usually require labels. The success of our method implies that leakage of intermediate representations may create a security breach for deployed models, which persists even when the model is transferred to downstream usage. Removal of batch norm weakens our attack, indicating it contributes to this vulnerability. Our attacks also succeed against LayerNorm empirically, thus being relevant for transformer architectures, most notably vision transformers which we analyze.

https://papers.nips.cc/paper_files/paper/2023/file/0266d95023740481d22d437aa8aba0e9-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/21559-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/0266d95023740481d22d437aa8aba0e9-Supplemental-Conference.zip

Detecting Any Human-Object Interaction Relationship: Universal HOI Detector with Spatial Prompt Learning on Foundation Models

Main Conference Track

Yichao Cao, Qingfei Tang, Xiu Su, Song Chen, Shan You, Xiaobo Lu, Chang Xu

Human-object interaction (HOI) detection aims to comprehend the intricate relationships between humans and objects, predicting triplets, and serving as the foundation for numerous computer vision tasks. The complexity and diversity of human-object interactions in the real world, however, pose significant challenges for both annotation and recognition, particularly in recognizing interactions within an open world context. This study explores the universal interaction recognition in an open-world setting through the use of Vision-Language (VL) foundation models and large language models (LLMs). The proposed method is dubbed as UniHOI. We conduct a deep analysis of the three hierarchical features inherent in visual HOI detectors and propose a method for high-level relation extraction aimed at VL foundation models, which we call HO prompt-based learning. Our design includes an HO Prompt-guided Decoder (HOPD), facilitates the association of high-level relation representations in the foundation model with various HO pairs within the image. Furthermore, we utilize a LLM (i.e. GPT) for interaction interpretation, generating a richer linguistic understanding for complex HOIs. For open-category interaction recognition, our method supports either of two input types: interaction phrase or interpretive sentence. Our efficient architecture design and learning methods effectively unleash the potential of the VL foundation models and LLMs, allowing UniHOI to surpass all existing methods with a substantial margin, under both supervised and zero-shot settings. The code and pre-trained weights will be made publicly available.

https://papers.nips.cc/paper_files/paper/2023/file/02687e7b22abc64e651be8da74ec610e-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/20272-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/02687e7b22abc64e651be8da74ec610e-Supplemental-Conference.pdf

Smoothing the Landscape Boosts the Signal for SGD: Optimal Sample Complexity for Learning Single Index Models

Main Conference Track

Alex Damian, Eshaan Nichani, Rong Ge, Jason D. Lee

We focus on the task of learning a single index model $\sigma(w^\star \cdot x)$ with respect to the isotropic Gaussian distribution in $d$ dimensions. Prior work has shown that the sample complexity of learning $w^\star$ is governed by the information exponent $k^\star$ of the link function $\sigma$, which is defined as the index of the first nonzero Hermite coefficient of $\sigma$. Ben Arous et al. (2021) showed that $n \gtrsim d^{k^\star-1}$ samples suffice for learning $w^\star$ and that this is tight for online SGD. However, the CSQ lower bound for gradient based methods only shows that $n \gtrsim d^{k^\star/2}$ samples are necessary. In this work, we close the gap between the upper and lower bounds by showing that online SGD on a smoothed loss learns $w^\star$ with $n \gtrsim d^{k^\star/2}$ samples. We also draw connections to statistical analyses of tensor PCA and to the implicit regularization effects of minibatch SGD on empirical losses.

https://papers.nips.cc/paper_files/paper/2023/file/02763667a5761ff92bb15d8751bcd223-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/19842-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/02763667a5761ff92bb15d8751bcd223-Supplemental-Conference.zip

A Scale-Invariant Sorting Criterion to Find a Causal Order in Additive Noise Models

Main Conference Track

Alexander Reisach, Myriam Tami, Christof Seiler, Antoine Chambaz, Sebastian Weichwald

Additive Noise Models (ANMs) are a common model class for causal discovery from observational data. Due to a lack of real-world data for which an underlying ANM is known, ANMs with randomly sampled parameters are commonly used to simulate data for the evaluation of causal discovery algorithms. While some parameters may be fixed by explicit assumptions, fully specifying an ANM requires choosing all parameters. Reisach et al. (2021) show that, for many ANM parameter choices, sorting the variables by increasing variance yields an ordering close to a causal order and introduce ‘var-sortability’ to quantify this alignment. Since increasing variances may be unrealistic and cannot be exploited when data scales are arbitrary, ANM data are often rescaled to unit variance in causal discovery benchmarking.We show that synthetic ANM data are characterized by another pattern that is scale-invariant and thus persists even after standardization: the explainable fraction of a variable’s variance, as captured by the coefficient of determination $R^2$, tends to increase along the causal order. The result is high ‘$R^2$-sortability’, meaning that sorting the variables by increasing $R^2$ yields an ordering close to a causal order. We propose a computationally efficient baseline algorithm termed ‘$R^2$-SortnRegress’ that exploits high $R^2$-sortability and that can match and exceed the performance of established causal discovery algorithms. We show analytically that sufficiently high edge weights lead to a relative decrease of the noise contributions along causal chains, resulting in increasingly deterministic relationships and high $R^2$. We characterize $R^2$-sortability on synthetic data with different simulation parameters and find high values in common settings. Our findings reveal high $R^2$-sortability as an assumption about the data generating process relevant to causal discovery and implicit in many ANM sampling schemes. It should be made explicit, as its prevalence in real-world data is an open question. For causal discovery benchmarking, we provide implementations of $R^2$-sortability, the $R^2$-SortnRegress algorithm, and ANM simulation procedures in our library CausalDisco at https://causaldisco.github.io/CausalDisco/.

https://papers.nips.cc/paper_files/paper/2023/file/027e86facfe7c1ea52ca1fca7bc1402b-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/19690-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/027e86facfe7c1ea52ca1fca7bc1402b-Supplemental-Conference.zip

PROTES: Probabilistic Optimization with Tensor Sampling

Main Conference Track

Anastasiia Batsheva, Andrei Chertkov, Gleb Ryzhakov, Ivan Oseledets

We developed a new method PROTES for black-box optimization, which is based on the probabilistic sampling from a probability density function given in the low-parametric tensor train format. We tested it on complex multidimensional arrays and discretized multivariable functions taken, among others, from real-world applications, including unconstrained binary optimization and optimal control problems, for which the possible number of elements is up to $2^{1000}$. In numerical experiments, both on analytic model functions and on complex problems, PROTES outperforms popular discrete optimization methods (Particle Swarm Optimization, Covariance Matrix Adaptation, Differential Evolution, and others).

https://papers.nips.cc/paper_files/paper/2023/file/028957869e560af14243ac37663a471e-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/21075-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/028957869e560af14243ac37663a471e-Supplemental-Conference.zip

Perturbation Towards Easy Samples Improves Targeted Adversarial Transferability

Main Conference Track

Junqi Gao, Biqing Qi, Yao Li, Zhichang Guo, Dong Li, Yuming Xing, Dazhi Zhang

The transferability of adversarial perturbations provides an effective shortcut for black-box attacks. Targeted perturbations have greater practicality but are more difficult to transfer between models. In this paper, we experimentally and theoretically demonstrated that neural networks trained on the same dataset have more consistent performance in High-Sample-Density-Regions (HSDR) of each class instead of low sample density regions. Therefore, in the target setting, adding perturbations towards HSDR of the target class is more effective in improving transferability. However, density estimation is challenging in high-dimensional scenarios. Further theoretical and experimental verification demonstrates that easy samples with low loss are more likely to be located in HSDR. Perturbations towards such easy samples in the target class can avoid density estimation for HSDR location. Based on the above facts, we verified that adding perturbations to easy samples in the target class improves targeted adversarial transferability of existing attack methods. A generative targeted attack strategy named Easy Sample Matching Attack (ESMA) is proposed, which has a higher success rate for targeted attacks and outperforms the SOTA generative method. Moreover, ESMA requires only $5\%$ of the storage space and much less computation time comparing to the current SOTA, as ESMA attacks all classes with only one model instead of seperate models for each class. Our code is available at https://github.com/gjq100/ESMA

https://papers.nips.cc/paper_files/paper/2023/file/028fcbcf85435d39a40c4d61b42c99a4-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/19716-/bibtex

AllSim: Simulating and Benchmarking Resource Allocation Policies in Multi-User Systems

Datasets and Benchmarks Track

Jeroen Berrevoets, Daniel Jarrett, Alex Chan, Mihaela van der Schaar

Numerous real-world systems, ranging from healthcare to energy grids, involve users competing for finite and potentially scarce resources. Designing policies for resource allocation in such real-world systems is challenging for many reasons, including the changing nature of user types and their (possibly urgent) need for resources. Researchers have developed numerous machine learning solutions for determining resource allocation policies in these challenging settings. However, a key limitation has been the absence of good methods and test-beds for benchmarking these policies; almost all resource allocation policies are benchmarked in environments which are either completely synthetic or do not allow any deviation from historical data. In this paper we introduce AllSim, which is a benchmarking environment for realistically simulating the impact and utility of policies for resource allocation in systems in which users compete for such scarce resources. Building such a benchmarking environment is challenging because it needs to successfully take into account the entire collective of potential users and the impact a resource allocation policy has on all the other users in the system. AllSim's benchmarking environment is modular (each component being parameterized individually), learnable (informed by historical data), and customizable (adaptable to changing conditions). These, when interacting with an allocation policy, produce a dataset of simulated outcomes for evaluation and comparison of such policies. We believe AllSim is an essential step towards a more systematic evaluation of policies for scarce resource allocation compared to current approaches for benchmarking such methods.

https://papers.nips.cc/paper_files/paper/2023/file/0296e17ec30fc36007edaaa2f96b5f17-Paper-Datasets_and_Benchmarks.pdf

https://papers.nips.cc/paper_files/paper/20181-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/0296e17ec30fc36007edaaa2f96b5f17-Supplemental-Datasets_and_Benchmarks.pdf

AVIS: Autonomous Visual Information Seeking with Large Language Model Agent

Main Conference Track

Ziniu Hu, Ahmet Iscen, Chen Sun, Kai-Wei Chang, Yizhou Sun, David Ross, Cordelia Schmid, Alireza Fathi

In this paper, we propose an autonomous information seeking visual question answering framework, AVIS. Our method leverages a Large Language Model (LLM) to dynamically strategize the utilization of external tools and to investigate their outputs via tree search, thereby acquiring the indispensable knowledge needed to provide answers to the posed questions. Responding to visual questions that necessitate external knowledge, such as "What event is commemorated by the building depicted in this image?", is a complex task. This task presents a combinatorial search space that demands a sequence of actions, including invoking APIs, analyzing their responses, and making informed decisions. We conduct a user study to collect a variety of instances of human decision-making when faced with this task. This data is then used to design a system comprised of three components: an LLM-powered planner that dynamically determines which tool to use next, an LLM-powered reasoner that analyzes and extracts key information from the tool outputs, and a working memory component that retains the acquired information throughout the process. The collected user behavior serves as a guide for our system in two key ways. First, we create a transition graph by analyzing the sequence of decisions made by users. This graph delineates distinct states and confines the set of actions available at each state. Second, we use examples of user decision-making to provide our LLM-powered planner and reasoner with relevant contextual instances, enhancing their capacity to make informed decisions. We show that AVIS achieves state-of-the-art results on knowledge-based visual question answering benchmarks such as Infoseek and OK-VQA.

https://papers.nips.cc/paper_files/paper/2023/file/029df12a9363313c3e41047844ecad94-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/19636-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/029df12a9363313c3e41047844ecad94-Supplemental-Conference.pdf

Conformal Prediction Sets for Ordinal Classification

Main Conference Track

Prasenjit Dey, Srujana Merugu, Sivaramakrishnan R Kaveri

Ordinal classification (OC), i.e., labeling instances along classes with a natural ordering, is common in multiple applications such as size or budget based recommendations and disease severity labeling. Often in practical scenarios, it is desirable to obtain a small set of likely classes with a guaranteed high chance of including the true class. Recent works on conformal prediction (CP) address this problem for the classification setting with non-ordered labels but the resulting prediction sets (PS) are often non-contiguous and unsuitable for ordinal classification. In this work, we propose a framework to adapt existing CP methods to generate contiguous sets with guaranteed coverage and minimal cardinality. Our framework employs a novel non-parametric approach for modeling unimodal distributions. Empirical results on both synthetic and real-world datasets demonstrate our method outperforms SOTA baselines by 4% on Accuracy@K and 8% on PS size.

https://papers.nips.cc/paper_files/paper/2023/file/029f699912bf3db747fe110948cc6169-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/21757-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/029f699912bf3db747fe110948cc6169-Supplemental-Conference.pdf

Minimax-Optimal Location Estimation

Main Conference Track

Shivam Gupta, Jasper Lee, Eric Price, Paul Valiant

Location estimation is one of the most basic questions in parametric statistics. Suppose we have a known distribution density $f$, and we get $n$ i.i.d. samples from $f(x-\mu)$ for some unknown shift $\mu$.The task is to estimate $\mu$ to high accuracy with high probability.The maximum likelihood estimator (MLE) is known to be asymptotically optimal as $n \to \infty$, but what is possible for finite $n$?In this paper, we give two location estimators that are optimal under different criteria: 1) an estimator that has minimax-optimal estimation error subject to succeeding with probability $1-\delta$ and 2) a confidence interval estimator which, subject to its output interval containing $\mu$ with probability at least $1-\delta$, has the minimum expected squared interval width among all shift-invariant estimators.The latter construction can be generalized to minimizing the expectation of any loss function on the interval width.

https://papers.nips.cc/paper_files/paper/2023/file/02a589ef9a4f6f1e2dcc1cfb3b978a51-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/22256-/bibtex

Tight Bounds for Volumetric Spanners and Applications

Main Conference Track

Aditya Bhaskara, Sepideh Mahabadi, Ali Vakilian

Given a set of points of interest, a volumetric spanner is a subset of the points using which all the points can be expressed using "small" coefficients (measured in an appropriate norm). Formally, given a set of vectors $X = [v_1, v_2, \dots, v_n]$, the goal is to find $T \subseteq [n]$ such that every $v \in X$ can be expressed as $\sum_{i\in T} \alpha_i v_i$, with $\Vert \alpha \Vert$ being small. This notion, which has also been referred to as a well-conditioned basis, has found several applications, including bandit linear optimization, determinant maximization, and matrix low rank approximation. In this paper, we give almost optimal bounds on the size of volumetric spanners for all $\ell_p$ norms, and show that they can be constructed using a simple local search procedure. We then show the applications of our result to other tasks and in particular the problem of finding coresets for the Minimum Volume Enclosing Ellipsoid (MVEE) problem.

https://papers.nips.cc/paper_files/paper/2023/file/02a92b52670752daf17b53f04f1ab405-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/19739-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/02a92b52670752daf17b53f04f1ab405-Supplemental-Conference.pdf

Wyze Rule: Federated Rule Dataset for Rule Recommendation Benchmarking

Datasets and Benchmarks Track

Mohammad Mahdi Kamani, Yuhang Yao, Hanjia Lyu, Zhongwei Cheng, Lin Chen, Liangju Li, Carlee Joe-Wong, Jiebo Luo

In the rapidly evolving landscape of smart home automation, the potential of IoT devices is vast. In this realm, rules are the main tool utilized for this automation, which are predefined conditions or triggers that establish connections between devices, enabling seamless automation of specific processes. However, one significant challenge researchers face is the lack of comprehensive datasets to explore and advance the field of smart home rule recommendations. These datasets are essential for developing and evaluating intelligent algorithms that can effectively recommend rules for automating processes while preserving the privacy of the users, as it involves personal information about users' daily lives. To bridge this gap, we present the Wyze Rule Dataset, a large-scale dataset designed specifically for smart home rule recommendation research. Wyze Rule encompasses over 1 million rules gathered from a diverse user base of 300,000 individuals from Wyze Labs, offering an extensive and varied collection of real-world data. With a focus on federated learning, our dataset is tailored to address the unique challenges of a cross-device federated learning setting in the recommendation domain, featuring a large-scale number of clients with widely heterogeneous data. To establish a benchmark for comparison and evaluation, we have meticulously implemented multiple baselines in both centralized and federated settings. Researchers can leverage these baselines to gauge the performance and effectiveness of their rule recommendation systems, driving advancements in the domain. The Wyze Rule Dataset is publicly accessible through HuggingFace's dataset API.

https://papers.nips.cc/paper_files/paper/2023/file/02b9d1e6d1b5295a6f883969ddc1bbbd-Paper-Datasets_and_Benchmarks.pdf

https://papers.nips.cc/paper_files/paper/21966-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/02b9d1e6d1b5295a6f883969ddc1bbbd-Supplemental-Datasets_and_Benchmarks.zip

Learning better with Dale’s Law: A Spectral Perspective

Main Conference Track

Pingsheng Li, Jonathan Cornford, Arna Ghosh, Blake Richards

Most recurrent neural networks (RNNs) do not include a fundamental constraint of real neural circuits: Dale's Law, which implies that neurons must be excitatory (E) or inhibitory (I). Dale's Law is generally absent from RNNs because simply partitioning a standard network's units into E and I populations impairs learning. However, here we extend a recent feedforward bio-inspired EI network architecture, named Dale's ANNs, to recurrent networks, and demonstrate that good performance is possible while respecting Dale's Law. This begs the question: What makes some forms of EI network learn poorly and others learn well? And, why does the simple approach of incorporating Dale's Law impair learning? Historically the answer was thought to be the sign constraints on EI network parameters, and this was a motivation behind Dale's ANNs. However, here we show the spectral properties of the recurrent weight matrix at initialisation are more impactful on network performance than sign constraints. We find that simple EI partitioning results in a singular value distribution that is multimodal and dispersed, whereas standard RNNs have an unimodal, more clustered singular value distribution, as do recurrent Dale's ANNs. We also show that the spectral properties and performance of partitioned EI networks are worse for small networks with fewer I units, and we present normalised SVD entropy as a measure of spectrum pathology that correlates with performance. Overall, this work sheds light on a long-standing mystery in neuroscience-inspired AI and computational neuroscience, paving the way for greater alignment between neural networks and biology.

https://papers.nips.cc/paper_files/paper/2023/file/02dd0db10c40092de3d9ec2508d12f60-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/22650-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/02dd0db10c40092de3d9ec2508d12f60-Supplemental-Conference.pdf

Dense-Exponential Random Features: Sharp Positive Estimators of the Gaussian Kernel

Main Conference Track

Valerii Likhosherstov, Krzysztof M Choromanski, Kumar Avinava Dubey, Frederick Liu, Tamas Sarlos, Adrian Weller

The problem of efficient approximation of a linear operator induced by the Gaussian or softmax kernel is often addressed using random features (RFs) which yield an unbiased approximation of the operator's result. Such operators emerge in important applications ranging from kernel methods to efficient Transformers. We propose parameterized, positive, non-trigonometric RFs which approximate Gaussian and softmax-kernels. In contrast to traditional RF approximations, parameters of these new methods can be optimized to reduce the variance of the approximation, and the optimum can be expressed in closed form. We show that our methods lead to variance reduction in practice (e^{10}-times smaller variance and beyond) and outperform previous methods in a kernel regression task. Using our proposed mechanism, we also present FAVOR#, a method for self-attention approximation in Transformers. We show that FAVOR# outperforms other random feature methods in speech modelling and natural language processing.

https://papers.nips.cc/paper_files/paper/2023/file/02dec8877fb7c6aa9a79f81661baca7c-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/20164-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/02dec8877fb7c6aa9a79f81661baca7c-Supplemental-Conference.zip

Projection-Free Online Convex Optimization via Efficient Newton Iterations

Main Conference Track

Khashayar Gatmiry, Zak Mhammedi

This paper presents new projection-free algorithms for Online Convex Optimization (OCO) over a convex domain $\mathcal{K} \subset \mathbb{R}^d$. Classical OCO algorithms (such as Online Gradient Descent) typically need to perform Euclidean projections onto the convex set $\mathcal{K}$ to ensure feasibility of their iterates. Alternative algorithms, such as those based on the Frank-Wolfe method, swap potentially-expensive Euclidean projections onto $\mathcal{K}$ for linear optimization over $\mathcal{K}$. However, such algorithms have a sub-optimal regret in OCO compared to projection-based algorithms. In this paper, we look at a third type of algorithms that output approximate Newton iterates using a self-concordant barrier for the set of interest. The use of a self-concordant barrier automatically ensures feasibility without the need of projections. However, the computation of the Newton iterates requires a matrix inverse, which can still be expensive. As our main contribution, we show how the stability of the Newton iterates can be leveraged to only compute the inverse Hessian a vanishing fractions of the rounds, leading to a new efficient projection-free OCO algorithm with a state-of-the-art regret bound.

https://papers.nips.cc/paper_files/paper/2023/file/03261886741f1f21f52f2a2d570616a2-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/22091-/bibtex

Read and Reap the Rewards: Learning to Play Atari with the Help of Instruction Manuals

Main Conference Track

Yue Wu, Yewen Fan, Paul Pu Liang, Amos Azaria, Yuanzhi Li, Tom M. Mitchell

High sample complexity has long been a challenge for RL. On the other hand, humans learn to perform tasks not only from interaction or demonstrations, but also by reading unstructured text documents, e.g., instruction manuals. Instruction manuals and wiki pages are among the most abundant data that could inform agents of valuable features and policies or task-specific environmental dynamics and reward structures. Therefore, we hypothesize that the ability to utilize human-written instruction manuals to assist learning policies for specific tasks should lead to a more efficient and better-performing agent. We propose the Read and Reward framework. Read and Reward speeds up RL algorithms on Atari games by reading manuals released by the Atari game developers. Our framework consists of a QA Extraction module that extracts and summarizes relevant information from the manual and a Reasoning module that evaluates object-agent interactions based on information from the manual. An auxiliary reward is then provided to a standard A2C RL agent, when interaction is detected. Experimentally, various RL algorithms obtain significant improvement in performance and training speed when assisted by our design. Code at github.com/Holmeswww/RnR

https://papers.nips.cc/paper_files/paper/2023/file/034d7bfeace2a9a258648b16fc626298-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/20787-/bibtex

Sharpness Minimization Algorithms Do Not Only Minimize Sharpness To Achieve Better Generalization

Main Conference Track

Kaiyue Wen, Zhiyuan Li, Tengyu Ma

Despite extensive studies, the underlying reason as to why overparameterizedneural networks can generalize remains elusive. Existing theory shows that common stochastic optimizers prefer flatter minimizers of the training loss, and thusa natural potential explanation is that flatness implies generalization. This workcritically examines this explanation. Through theoretical and empirical investigation, we identify the following three scenarios for two-layer ReLU networks: (1)flatness provably implies generalization; (2) there exist non-generalizing flattestmodels and sharpness minimization algorithms fail to generalize poorly, and (3)perhaps most strikingly, there exist non-generalizing flattest models, but sharpnessminimization algorithms still generalize. Our results suggest that the relationshipbetween sharpness and generalization subtly depends on the data distributionsand the model architectures and sharpness minimization algorithms do not onlyminimize sharpness to achieve better generalization. This calls for the search forother explanations for the generalization of over-parameterized neural networks

https://papers.nips.cc/paper_files/paper/2023/file/0354767c6386386be17cabe4fc59711b-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/21224-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/0354767c6386386be17cabe4fc59711b-Supplemental-Conference.zip

Feature-Learning Networks Are Consistent Across Widths At Realistic Scales

Main Conference Track

Nikhil Vyas, Alexander Atanasov, Blake Bordelon, Depen Morwani, Sabarish Sainathan, Cengiz Pehlevan

We study the effect of width on the dynamics of feature-learning neural networks across a variety of architectures and datasets. Early in training, wide neural networks trained on online data have not only identical loss curves but also agree in their point-wise test predictions throughout training. For simple tasks such as CIFAR-5m this holds throughout training for networks of realistic widths. We also show that structural properties of the models, including internal representations, preactivation distributions, edge of stability phenomena, and large learning rate effects are consistent across large widths. This motivates the hypothesis that phenomena seen in realistic models can be captured by infinite-width, feature-learning limits. For harder tasks (such as ImageNet and language modeling), and later training times, finite-width deviations grow systematically. Two distinct effects cause these deviations across widths. First, the network output has an initialization-dependent variance scaling inversely with width, which can be removed by ensembling networks. We observe, however, that ensembles of narrower networks perform worse than a single wide network. We call this the bias of narrower width. We conclude with a spectral perspective on the origin of this finite-width bias.

https://papers.nips.cc/paper_files/paper/2023/file/03600ae6c3392fd65ad7c3a90c6f7ce8-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/22545-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/03600ae6c3392fd65ad7c3a90c6f7ce8-Supplemental-Conference.zip

Taylor TD-learning

Main Conference Track

Michele Garibbo, Maxime Robeyns, Laurence Aitchison

Many reinforcement learning approaches rely on temporal-difference (TD) learning to learn a critic.However, TD-learning updates can be high variance.Here, we introduce a model-based RL framework, Taylor TD, which reduces this variance in continuous state-action settings. Taylor TD uses a first-order Taylor series expansion of TD updates.This expansion allows Taylor TD to analytically integrate over stochasticity in the action-choice, and some stochasticity in the state distribution for the initial state and action of each TD update.We include theoretical and empirical evidence that Taylor TD updates are indeed lower variance than standard TD updates. Additionally, we show Taylor TD has the same stable learning guarantees as standard TD-learning with linear function approximation under a reasonable assumption.Next, we combine Taylor TD with the TD3 algorithm, forming TaTD3.We show TaTD3 performs as well, if not better, than several state-of-the art model-free and model-based baseline algorithms on a set of standard benchmark tasks.

https://papers.nips.cc/paper_files/paper/2023/file/036912a83bdbb1fd792baf6532f102d8-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/22597-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/036912a83bdbb1fd792baf6532f102d8-Supplemental-Conference.pdf

Calibrating Neural Simulation-Based Inference with Differentiable Coverage Probability

Main Conference Track

Maciej Falkiewicz, Naoya Takeishi, Imahn Shekhzadeh, Antoine Wehenkel, Arnaud Delaunoy, Gilles Louppe, Alexandros Kalousis

Bayesian inference allows expressing the uncertainty of posterior belief under a probabilistic model given prior information and the likelihood of the evidence. Predominantly, the likelihood function is only implicitly established by a simulator posing the need for simulation-based inference (SBI). However, the existing algorithms can yield overconfident posteriors (Hermans et al., 2022) defeating the whole purpose of credibility if the uncertainty quantification is inaccurate. We propose to include a calibration term directly into the training objective of the neural model in selected amortized SBI techniques. By introducing a relaxation of the classical formulation of calibration error we enable end-to-end backpropagation. The proposed method is not tied to any particular neural model and brings moderate computational overhead compared to the profits it introduces. It is directly applicable to existing computational pipelines allowing reliable black-box posterior inference. We empirically show on six benchmark problems that the proposed method achieves competitive or better results in terms of coverage and expected posterior density than the previously existing approaches.

https://papers.nips.cc/paper_files/paper/2023/file/03a9a9c1e15850439653bb971a4ad4b3-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/21076-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/03a9a9c1e15850439653bb971a4ad4b3-Supplemental-Conference.zip

Agnostic Multi-Group Active Learning

Main Conference Track

Nicholas Rittler, Kamalika Chaudhuri

Inspired by the problem of improving classification accuracy on rare or hard subsets of a population, there has been recent interest in models of learning where the goal is to generalize to a collection of distributions, each representing a ``group''. We consider a variant of this problem from the perspective of active learning, where the learner is endowed with the power to decide which examples are labeled from each distribution in the collection, and the goal is to minimize the number of label queries while maintaining PAC-learning guarantees. Our main challenge is that standard active learning techniques such as disagreement-based active learning do not directly apply to the multi-group learning objective. We modify existing algorithms to provide a consistent active learning algorithm for an agnostic formulation of multi-group learning, which given a collection of $G$ distributions and a hypothesis class $\mathcal{H}$ with VC-dimension $d$, outputs an $\epsilon$-optimal hypothesis using $\tilde{O}\left( (\nu^2/\epsilon^2) G d \theta_{\mathcal{G}}^2 \log^2(1/\epsilon) + G\log(1/\epsilon)/\epsilon^2 \right)$ label queries, where $\theta_{\mathcal{G}}$ is the worst-case disagreement coefficient over the collection. Roughly speaking, this guarantee improves upon the label complexity of standard multi-group learning in regimes where disagreement-based active learning algorithms may be expected to succeed, and the number of groups is not too large. We also consider the special case where each distribution in the collection is individually realizable with respect to $\mathcal{H}$, and demonstrate $\tilde{O}\left( G d \theta_{\mathcal{G}} \log(1/\epsilon) \right)$ label queries are sufficient for learning in this case. We further give an approximation result for the full agnostic case inspired by the group realizable strategy.

https://papers.nips.cc/paper_files/paper/2023/file/03b1043052700b1a471996b0baf309d4-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/22161-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/03b1043052700b1a471996b0baf309d4-Supplemental-Conference.pdf

Self-Weighted Contrastive Learning among Multiple Views for Mitigating Representation Degeneration

Main Conference Track

Jie Xu, Shuo Chen, Yazhou Ren, Xiaoshuang Shi, Hengtao Shen, Gang Niu, Xiaofeng Zhu

Recently, numerous studies have demonstrated the effectiveness of contrastive learning (CL), which learns feature representations by pulling in positive samples while pushing away negative samples. Many successes of CL lie in that there exists semantic consistency between data augmentations of the same instance. In multi-view scenarios, however, CL might cause representation degeneration when the collected multiple views inherently have inconsistent semantic information or their representations subsequently do not capture sufficient discriminative information. To address this issue, we propose a novel framework called SEM: SElf-weighted Multi-view contrastive learning with reconstruction regularization. Specifically, SEM is a general framework where we propose to first measure the discrepancy between pairwise representations and then minimize the corresponding self-weighted contrastive loss, and thus making SEM adaptively strengthen the useful pairwise views and also weaken the unreliable pairwise views. Meanwhile, we impose a self-supervised reconstruction term to regularize the hidden features of encoders, to assist CL in accessing sufficient discriminative information of data. Experiments on public multi-view datasets verified that SEM can mitigate representation degeneration in existing CL methods and help them achieve significant performance improvements. Ablation studies also demonstrated the effectiveness of SEM with different options of weighting strategies and reconstruction terms.

https://papers.nips.cc/paper_files/paper/2023/file/03b13b0db740b95cb741e007178ef5e5-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/20190-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/03b13b0db740b95cb741e007178ef5e5-Supplemental-Conference.pdf

Neural Polarizer: A Lightweight and Effective Backdoor Defense via Purifying Poisoned Features

Main Conference Track

Mingli Zhu, Shaokui Wei, Hongyuan Zha, Baoyuan Wu

Recent studies have demonstrated the susceptibility of deep neural networks to backdoor attacks. Given a backdoored model, its prediction of a poisoned sample with trigger will be dominated by the trigger information, though trigger information and benign information coexist. Inspired by the mechanism of the optical polarizer that a polarizer could pass light waves with particular polarizations while filtering light waves with other polarizations, we propose a novel backdoor defense method by inserting a learnable neural polarizer into the backdoored model as an intermediate layer, in order to purify the poisoned sample via filtering trigger information while maintaining benign information. The neural polarizer is instantiated as one lightweight linear transformation layer, which is learned through solving a well designed bi-level optimization problem, based on a limited clean dataset. Compared to other fine-tuning-based defense methods which often adjust all parameters of the backdoored model, the proposed method only needs to learn one additional layer, such that it is more efficient and requires less clean data. Extensive experiments demonstrate the effectiveness and efficiency of our method in removing backdoors across various neural network architectures and datasets, especially in the case of very limited clean data. Codes are available at \href{https://github.com/SCLBD/BackdoorBench}{https://github.com/SCLBD/BackdoorBench} (PyTorch) and \href{https://github.com/JulieCarlon/NPD-MindSpore}{https://github.com/JulieCarlon/NPD-MindSpore} (MindSpore).

https://papers.nips.cc/paper_files/paper/2023/file/03df5246cc78af497940338dd3eacbaa-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/22465-/bibtex

Tools for Verifying Neural Models' Training Data

Main Conference Track

Dami Choi, Yonadav Shavit, David K. Duvenaud

It is important that consumers and regulators can verify the provenance of large neural models to evaluate their capabilities and risks. We introduce the concept of a "Proof-of-Training-Data": any protocol that allows a model trainer to convince a Verifier of the training data that produced a set of model weights. Such protocols could verify the amount and kind of data and compute used to train the model, including whether it was trained on specific harmful or beneficial data sources. We explore efficient verification strategies for Proof-of-Training-Data that are compatible with most current large-model training procedures. These include a method for the model-trainer to verifiably pre-commit to a random seed used in training, and a method that exploits models' tendency to temporarily overfit to training data in order to detect whether a given data-point was included in training. We show experimentally that our verification procedures can catch a wide variety of attacks, including all known attacks from the Proof-of-Learning literature.

https://papers.nips.cc/paper_files/paper/2023/file/03e33e1f62e3302b47fe1d38a235921e-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/22555-/bibtex

Towards Higher Ranks via Adversarial Weight Pruning

Main Conference Track

Yuchuan Tian, Hanting Chen, Tianyu Guo, Chao Xu, Yunhe Wang

Convolutional Neural Networks (CNNs) are hard to deploy on edge devices due to its high computation and storage complexities. As a common practice for model compression, network pruning consists of two major categories: unstructured and structured pruning, where unstructured pruning constantly performs better. However, unstructured pruning presents a structured pattern at high pruning rates, which limits its performance. To this end, we propose a Rank-based PruninG (RPG) method to maintain the ranks of sparse weights in an adversarial manner. In each step, we minimize the low-rank approximation error for the weight matrices using singular value decomposition, and maximize their distance by pushing the weight matrices away from its low rank approximation. This rank-based optimization objective guides sparse weights towards a high-rank topology. The proposed method is conducted in a gradual pruning fashion to stabilize the change of rank during training. Experimental results on various datasets and different tasks demonstrate the effectiveness of our algorithm in high sparsity. The proposed RPG outperforms the state-of-the-art performance by 1.13\% top-1 accuracy on ImageNet in ResNet-50 with 98\% sparsity. The codes are available at https://github.com/huawei-noah/Efficient-Computing/tree/master/Pruning/RPG and https://gitee.com/mindspore/models/tree/master/research/cv/RPG.

https://papers.nips.cc/paper_files/paper/2023/file/040ace837dd270a87055bb10dd7c0392-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/21480-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/040ace837dd270a87055bb10dd7c0392-Supplemental-Conference.pdf

On the Overlooked Pitfalls of Weight Decay and How to Mitigate Them: A Gradient-Norm Perspective

Main Conference Track

Zeke Xie, Zhiqiang Xu, Jingzhao Zhang, Issei Sato, Masashi Sugiyama

Weight decay is a simple yet powerful regularization technique that has been very widely used in training of deep neural networks (DNNs). While weight decay has attracted much attention, previous studies fail to discover some overlooked pitfalls on large gradient norms resulted by weight decay. In this paper, we discover that, weight decay can unfortunately lead to large gradient norms at the final phase (or the terminated solution) of training, which often indicates bad convergence and poor generalization. To mitigate the gradient-norm-centered pitfalls, we present the first practical scheduler for weight decay, called the Scheduled Weight Decay (SWD) method that can dynamically adjust the weight decay strength according to the gradient norm and significantly penalize large gradient norms during training. Our experiments also support that SWD indeed mitigates large gradient norms and often significantly outperforms the conventional constant weight decay strategy for Adaptive Moment Estimation (Adam).

https://papers.nips.cc/paper_files/paper/2023/file/040d3b6af368bf71f952c18da5713b48-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/19692-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/040d3b6af368bf71f952c18da5713b48-Supplemental-Conference.pdf

Leveraging Early-Stage Robustness in Diffusion Models for Efficient and High-Quality Image Synthesis

Main Conference Track

Yulhwa Kim, Dongwon Jo, Hyesung Jeon, Taesu Kim, Daehyun Ahn, Hyungjun Kim, jae-joon kim

While diffusion models have demonstrated exceptional image generation capabilities, the iterative noise estimation process required for these models is compute-intensive and their practical implementation is limited by slow sampling speeds. In this paper, we propose a novel approach to speed up the noise estimation network by leveraging the robustness of early-stage diffusion models. Our findings indicate that inaccurate computation during the early-stage of the reverse diffusion process has minimal impact on the quality of generated images, as this stage primarily outlines the image while later stages handle the finer details that require more sensitive information. To improve computational efficiency, we combine our findings with post-training quantization (PTQ) to introduce a method that utilizes low-bit activation for the early reverse diffusion process while maintaining high-bit activation for the later stages. Experimental results show that the proposed method can accelerate the early-stage computation without sacrificing the quality of the generated images.

https://papers.nips.cc/paper_files/paper/2023/file/04261fce1705c4f02f062866717d592a-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/21515-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/04261fce1705c4f02f062866717d592a-Supplemental-Conference.zip

Adversarial Model for Offline Reinforcement Learning

Main Conference Track

Mohak Bhardwaj, Tengyang Xie, Byron Boots, Nan Jiang, Ching-An Cheng

We propose a novel model-based offline Reinforcement Learning (RL) framework, called Adversarial Model for Offline Reinforcement Learning (ARMOR), which can robustly learn policies to improve upon an arbitrary reference policy regardless of data coverage. ARMOR is designed to optimize policies for the worst-case performance relative to the reference policy through adversarially training a Markov decision process model. In theory, we prove that ARMOR, with a well-tuned hyperparameter, can compete with the best policy within data coverage when the reference policy is supported by the data. At the same time, ARMOR is robust to hyperparameter choices: the policy learned by ARMOR, with any admissible hyperparameter, would never degrade the performance of the reference policy, even when the reference policy is not covered by the dataset. To validate these properties in practice, we design a scalable implementation of ARMOR, which by adversarial training, can optimize policies without using model ensembles in contrast to typical model-based methods. We show that ARMOR achieves competent performance with both state-of-the-art offline model-free and model-based RL algorithms and can robustly improve the reference policy over various hyperparameter choices.

https://papers.nips.cc/paper_files/paper/2023/file/0429ececfb199efc93182990169e73bb-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/19741-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/0429ececfb199efc93182990169e73bb-Supplemental-Conference.zip

Training Your Image Restoration Network Better with Random Weight Network as Optimization Function

Main Conference Track

man zhou, Naishan Zheng, Yuan Xu, Chun-Le Guo, Chongyi Li

The blooming progress made in deep learning-based image restoration has been largely attributed to the availability of high-quality, large-scale datasets and advanced network structures. However, optimization functions such as L1 and L2 are still de facto. In this study, we propose to investigate new optimization functions to improve image restoration performance. Our key insight is that ``random weight network can be acted as a constraint for training better image restoration networks''. However, not all random weight networks are suitable as constraints. We draw inspiration from Functional theory and show that alternative random weight networks should be represented in the form of a strict mathematical manifold. We explore the potential of our random weight network prototypes that satisfy this requirement: Taylor's unfolding network, invertible neural network, central difference convolution, and zero-order filtering. We investigate these prototypes from four aspects: 1) random weight strategies, 2) network architectures, 3) network depths, and 4) combinations of random weight networks. Furthermore, we devise the random weight in two variants: the weights are randomly initialized only once during the entire training procedure, and the weights are randomly initialized in each training epoch. Our approach can be directly integrated into existing networks without incurring additional training and testing computational costs. We perform extensive experiments across multiple image restoration tasks, including image denoising, low-light image enhancement, and guided image super-resolution to demonstrate the consistent performance gains achieved by our method. Upon acceptance of this paper, we will release the code.

https://papers.nips.cc/paper_files/paper/2023/file/043f0503c4f652c737add3690aa5d12c-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/20909-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/043f0503c4f652c737add3690aa5d12c-Supplemental-Conference.pdf

Passive learning of active causal strategies in agents and language models

Main Conference Track

Andrew Lampinen, Stephanie Chan, Ishita Dasgupta, Andrew Nam, Jane Wang

What can be learned about causality and experimentation from passive data? This question is salient given recent successes of passively-trained language models in interactive domains such as tool use. Passive learning is inherently limited. However, we show that purely passive learning can in fact allow an agent to learn generalizable strategies for determining and using causal structures, as long as the agent can intervene at test time. We formally illustrate that learning a strategy of first experimenting, then seeking goals, can allow generalization from passive learning in principle. We then show empirically that agents trained via imitation on expert data can indeed generalize at test time to infer and use causal links which are never present in the training data; these agents can also generalize experimentation strategies to novel variable sets never observed in training.We then show that strategies for causal intervention and exploitation can be generalized from passive data even in a more complex environment with high-dimensional observations, with the support of natural language explanations. Explanations can even allow passive learners to generalize out-of-distribution from perfectly-confounded training data. Finally, we show that language models, trained only on passive next-word prediction, can generalize causal intervention strategies from a few-shot prompt containing explanations and reasoning. These results highlight the surprising power of passive learning of active causal strategies, and have implications for understanding the behaviors and capabilities of language models.

https://papers.nips.cc/paper_files/paper/2023/file/045c87def0c02e3ad0d3d849766d7f1e-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/22193-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/045c87def0c02e3ad0d3d849766d7f1e-Supplemental-Conference.pdf

Zero-Regret Performative Prediction Under Inequality Constraints

Main Conference Track

Wenjing YAN, Xuanyu Cao

Performative prediction is a recently proposed framework where predictions guide decision-making and hence influence future data distributions. Such performative phenomena are ubiquitous in various areas, such as transportation, finance, public policy, and recommendation systems. To date, work on performative prediction has only focused on unconstrained problems, neglecting the fact that many real-world learning problems are subject to constraints. This paper bridges this gap by studying performative prediction under inequality constraints. Unlike most existing work that provides only performative stable points, we aim to find the optimal solutions. Anticipating performative gradient is a challenging task, due to the agnostic performative effect on data distributions. To address this issue, we first develop a robust primal-dual framework that requires only approximate gradients up to a certain accuracy, yet delivers the same order of performance as the stationary stochastic primal-dual algorithm without performativity. Based on this framework, we then propose an adaptive primal-dual algorithm for location families. Our analysis demonstrates that the proposed adaptive primal-dual algorithm attains $\mathcal{O}(\sqrt{T})$ regret and constraint violations, using only $\sqrt{T} + 2T$ samples, where $T$ is the time horizon. To our best knowledge, this is the first study and analysis on the optimality of the performative prediction problem under inequality constraints. Finally, we validate the effectiveness of our algorithm and theoretical results through numerical simulations.

https://papers.nips.cc/paper_files/paper/2023/file/047397849f63b4fcfced4ff720159f3d-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/20204-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/047397849f63b4fcfced4ff720159f3d-Supplemental-Conference.pdf

Towards Free Data Selection with General-Purpose Models

Main Conference Track

Yichen Xie, Mingyu Ding, Masayoshi TOMIZUKA, Wei Zhan

A desirable data selection algorithm can efficiently choose the most informative samples to maximize the utility of limited annotation budgets. However, current approaches, represented by active learning methods, typically follow a cumbersome pipeline that iterates the time-consuming model training and batch data selection repeatedly. In this paper, we challenge this status quo by designing a distinct data selection pipeline that utilizes existing general-purpose models to select data from various datasets with a single-pass inference without the need for additional training or supervision. A novel free data selection (FreeSel) method is proposed following this new pipeline. Specifically, we define semantic patterns extracted from inter-mediate features of the general-purpose model to capture subtle local information in each image. We then enable the selection of all data samples in a single pass through distance-based sampling at the fine-grained semantic pattern level. FreeSel bypasses the heavy batch selection process, achieving a significant improvement in efficiency and being 530x faster than existing active learning methods. Extensive experiments verify the effectiveness of FreeSel on various computer vision tasks.

https://papers.nips.cc/paper_files/paper/2023/file/047682108c3b053c61ad2da5a6057b4e-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/21335-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/047682108c3b053c61ad2da5a6057b4e-Supplemental-Conference.pdf

Communication-Efficient Federated Bilevel Optimization with Global and Local Lower Level Problems

Main Conference Track

Junyi Li, Feihu Huang, Heng Huang

Bilevel Optimization has witnessed notable progress recently with new emerging efficient algorithms. However, its application in the Federated Learning setting remains relatively underexplored, and the impact of Federated Learning's inherent challenges on the convergence of bilevel algorithms remain obscure.In this work, we investigate Federated Bilevel Optimization problems and propose a communication-efficient algorithm, named FedBiOAcc. The algorithm leverages an efficient estimation of the hyper-gradient in the distributed setting and utilizes the momentum-based variance-reduction acceleration. Remarkably, FedBiOAcc achieves a communication complexity $O(\epsilon^{-1})$, a sample complexity $O(\epsilon^{-1.5})$ and the linear speed up with respect to the number of clients. We also analyze a special case of the Federated Bilevel Optimization problems, where lower level problems are locally managed by clients. We prove that FedBiOAcc-Local, a modified version of FedBiOAcc, converges at the same rate for this type of problems. Finally, we validate the proposed algorithms through two real-world tasks: Federated Data-cleaning and Federated Hyper-representation Learning. Empirical results show superior performance of our algorithms.

https://papers.nips.cc/paper_files/paper/2023/file/04bd683d5428d91c5fbb5a7d2c27064d-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/20233-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/04bd683d5428d91c5fbb5a7d2c27064d-Supplemental-Conference.pdf

Partial Multi-Label Learning with Probabilistic Graphical Disambiguation

Main Conference Track

Jun-Yi Hang, Min-Ling Zhang

In partial multi-label learning (PML), each training example is associated with a set of candidate labels, among which only some labels are valid. As a common strategy to tackle PML problem, disambiguation aims to recover the ground-truth labeling information from such inaccurate annotations. However, existing approaches mainly rely on heuristics or ad-hoc rules to disambiguate candidate labels, which may not be universal enough in complicated real-world scenarios. To provide a principled way for disambiguation, we make a first attempt to explore the probabilistic graphical model for PML problem, where a directed graph is tailored to infer latent ground-truth labeling information from the generative process of partial multi-label data. Under the framework of stochastic gradient variational Bayes, a unified variational lower bound is derived for this graphical model, which is further relaxed probabilistically so that the desired prediction model can be induced with simultaneously identified ground-truth labeling information. Comprehensive experiments on multiple synthetic and real-world data sets show that our approach outperforms the state-of-the-art counterparts.

https://papers.nips.cc/paper_files/paper/2023/file/04e05ba5cbc36044f6499d1edf15247e-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/21623-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/04e05ba5cbc36044f6499d1edf15247e-Supplemental-Conference.pdf

Reward Scale Robustness for Proximal Policy Optimization via DreamerV3 Tricks

Main Conference Track

Ryan Sullivan, Akarsh Kumar, Shengyi Huang, John Dickerson, Joseph Suarez

Most reinforcement learning methods rely heavily on dense, well-normalized environment rewards. DreamerV3 recently introduced a model-based method with a number of tricks that mitigate these limitations, achieving state-of-the-art on a wide range of benchmarks with a single set of hyperparameters. This result sparked discussion about the generality of the tricks, since they appear to be applicable to other reinforcement learning algorithms. Our work applies DreamerV3's tricks to PPO and is the first such empirical study outside of the original work. Surprisingly, we find that the tricks presented do not transfer as general improvements to PPO. We use a high quality PPO reference implementation and present extensive ablation studies totaling over 10,000 A100 hours on the Arcade Learning Environment and the DeepMind Control Suite. Though our experiments demonstrate that these tricks do not generally outperform PPO, we identify cases where they succeed and offer insight into the relationship between the implementation tricks. In particular, PPO with these tricks performs comparably to PPO on Atari games with reward clipping and significantly outperforms PPO without reward clipping.

https://papers.nips.cc/paper_files/paper/2023/file/04f61ec02d1b3a025a59d978269ce437-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/19965-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/04f61ec02d1b3a025a59d978269ce437-Supplemental-Conference.pdf

Emergent Correspondence from Image Diffusion

Main Conference Track

Luming Tang, Menglin Jia, Qianqian Wang, Cheng Perng Phoo, Bharath Hariharan

Finding correspondences between images is a fundamental problem in computer vision. In this paper, we show that correspondence emerges in image diffusion models without any explicit supervision. We propose a simple strategy to extract this implicit knowledge out of diffusion networks as image features, namely DIffusion FeaTures (DIFT), and use them to establish correspondences between real images. Without any additional fine-tuning or supervision on the task-specific data or annotations, DIFT is able to outperform both weakly-supervised methods and competitive off-the-shelf features in identifying semantic, geometric, and temporal correspondences. Particularly for semantic correspondence, DIFT from Stable Diffusion is able to outperform DINO and OpenCLIP by 19 and 14 accuracy points respectively on the challenging SPair-71k benchmark. It even outperforms the state-of-the-art supervised methods on 9 out of 18 categories while remaining on par for the overall performance. Project page: https://diffusionfeatures.github.io.

https://papers.nips.cc/paper_files/paper/2023/file/0503f5dce343a1d06d16ba103dd52db1-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/19892-/bibtex

Robust Learning with Progressive Data Expansion Against Spurious Correlation

Main Conference Track

Yihe Deng, Yu Yang, Baharan Mirzasoleiman, Quanquan Gu

While deep learning models have shown remarkable performance in various tasks, they are susceptible to learning non-generalizable _spurious features_ rather than the core features that are genuinely correlated to the true label. In this paper, beyond existing analyses of linear models, we theoretically examine the learning process of a two-layer nonlinear convolutional neural network in the presence of spurious features. Our analysis suggests that imbalanced data groups and easily learnable spurious features can lead to the dominance of spurious features during the learning process. In light of this, we propose a new training algorithm called **PDE** that efficiently enhances the model's robustness for a better worst-group performance. PDE begins with a group-balanced subset of training data and progressively expands it to facilitate the learning of the core features. Experiments on synthetic and real-world benchmark datasets confirm the superior performance of our method on models such as ResNets and Transformers. On average, our method achieves a $2.8$ \% improvement in worst-group accuracy compared with the state-of-the-art method, while enjoying up to $10\times$ faster training efficiency.

https://papers.nips.cc/paper_files/paper/2023/file/0506ad3d1bcc8398a920db9340f27fe4-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/22024-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/0506ad3d1bcc8398a920db9340f27fe4-Supplemental-Conference.pdf

Multiclass Boosting: Simple and Intuitive Weak Learning Criteria

Main Conference Track

Nataly Brukhim, Amit Daniely, Yishay Mansour, Shay Moran

We study a generalization of boosting to the multiclass setting.We introduce a weak learning condition for multiclass classification that captures the original notion of weak learnability as being “slightly better than random guessing”. We give a simple and efficient boosting algorithm, that does not require realizability assumptions and its sample and oracle complexity bounds are independent of the number of classes. In addition, we utilize our new boosting technique in several theoretical applications within the context of List PAC Learning. First, we establish an equivalence to weak PAC learning. Furthermore, we present a new result on boosting for list learners, as well as provide a novel proof for the characterization of multiclass PAC learning and List PAC learning. Notably, our technique gives rise to simplified algorithms and analysis compared to previous works.

https://papers.nips.cc/paper_files/paper/2023/file/050f8591be3874b52fdac4e1060eeb29-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/20243-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/050f8591be3874b52fdac4e1060eeb29-Supplemental-Conference.pdf

Approximate Heavy Tails in Offline (Multi-Pass) Stochastic Gradient Descent

Main Conference Track

Kruno Lehman, Alain Durmus, Umut Simsekli

A recent line of empirical studies has demonstrated that SGD might exhibit a heavy-tailed behavior in practical settings, and the heaviness of the tails might correlate with the overall performance. In this paper, we investigate the emergence of such heavy tails. Previous works on this problem only considered, up to our knowledge, online (also called single-pass) SGD, in which the emergence of heavy tails in theoretical findings is contingent upon access to an infinite amount of data. Hence, the underlying mechanism generating the reported heavy-tailed behavior in practical settings, where the amount of training data is finite, is still not well-understood. Our contribution aims to fill this gap. In particular, we show that the stationary distribution of offline (also called multi-pass) SGD exhibits ‘approximate’ power-law tails and the approximation error is controlled by how fast the empirical distribution of the training data converges to the true underlying data distribution in the Wasserstein metric. Our main takeaway is that, as the number of data points increases, offline SGD will behave increasingly ‘power-law-like’. To achieve this result, we first prove nonasymptotic Wasserstein convergence bounds for offline SGD to online SGD as the number of data points increases, which can be interesting on their own. Finally, we illustrate our theory on various experiments conducted on synthetic data and neural networks.

https://papers.nips.cc/paper_files/paper/2023/file/0525a72df7fb2cd943c780d059b94774-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/19638-/bibtex

Uncovering Neural Scaling Laws in Molecular Representation Learning

Datasets and Benchmarks Track

Dingshuo Chen, Yanqiao Zhu, Jieyu Zhang, Yuanqi Du, Zhixun Li, Qiang Liu, Shu Wu, Liang Wang

Molecular Representation Learning (MRL) has emerged as a powerful tool for drug and materials discovery in a variety of tasks such as virtual screening and inverse design. While there has been a surge of interest in advancing model-centric techniques, the influence of both data quantity and quality on molecular representations is not yet clearly understood within this field. In this paper, we delve into the neural scaling behaviors of MRL from a data-centric viewpoint, examining four key dimensions: (1) data modalities, (2) dataset splitting, (3) the role of pre-training, and (4) model capacity.Our empirical studies confirm a consistent power-law relationship between data volume and MRL performance across these dimensions. Additionally, through detailed analysis, we identify potential avenues for improving learning efficiency.To challenge these scaling laws, we adapt seven popular data pruning strategies to molecular data and benchmark their performance. Our findings underline the importance of data-centric MRL and highlight possible directions for future research.

https://papers.nips.cc/paper_files/paper/2023/file/052e22cfdd344c79634f7ec76fa03e22-Paper-Datasets_and_Benchmarks.pdf

https://papers.nips.cc/paper_files/paper/19545-/bibtex

FlowCam: Training Generalizable 3D Radiance Fields without Camera Poses via Pixel-Aligned Scene Flow

Main Conference Track

Cameron Smith, Yilun Du, Ayush Tewari, Vincent Sitzmann

Reconstruction of 3D neural fields from posed images has emerged as a promising method for self-supervised representation learning. The key challenge preventing the deployment of these 3D scene learners on large-scale video data is their dependence on precise camera poses from structure-from-motion, which is prohibitively expensive to run at scale. We propose a method that jointly reconstructs camera poses and 3D neural scene representations online and in a single forward pass. We estimate poses by first lifting frame-to-frame optical flow to 3D scene flow via differentiable rendering, preserving locality and shift-equivariance of the image processing backbone. SE(3) camera pose estimation is then performed via a weighted least-squares fit to the scene flow field. This formulation enables us to jointly supervise pose estimation and a generalizable neural scene representation via re-rendering the input video, and thus, train end-to-end and fully self-supervised on real-world video datasets. We demonstrate that our method performs robustly on diverse, real-world video, notably on sequences traditionally challenging to optimization-based pose estimation techniques.

https://papers.nips.cc/paper_files/paper/2023/file/0534abc9e6db91683d82186ef0d68202-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/21853-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/0534abc9e6db91683d82186ef0d68202-Supplemental-Conference.zip

Minimum Description Length and Generalization Guarantees for Representation Learning

Main Conference Track

Milad Sefidgaran, Abdellatif Zaidi, Piotr Krasnowski

A major challenge in designing efficient statistical supervised learning algorithms is finding representations that perform well not only on available training samples but also on unseen data. While the study of representation learning has spurred much interest, most existing such approaches are heuristic; and very little is known about theoretical generalization guarantees. For example, the information bottleneck method seeks a good generalization by finding a minimal description of the input that is maximally informative about the label variable, where minimality and informativeness are both measured by Shannon’s mutual information. In this paper, we establish a compressibility framework that allows us to derive upper bounds on the generalization error of a representation learning algorithm in terms of the ``Minimum Description Length'' (MDL) of the labels or the latent variables (representations). Rather than the mutual information between the encoder’s input and the representation, which is often believed to reflect the algorithm’s generalization capability in the related literature but in fact, falls short of doing so, our new bounds involve the "multi-letter" relative entropy between the distribution of the representations (or labels) of the training and test sets and a fixed prior. In particular, these new bounds reflect the structure of the encoder and are not vacuous for deterministic algorithms. Our compressibility approach, which is information-theoretic in nature, builds upon that of Blum-Langford for PAC-MDL bounds and introduces two essential ingredients: block-coding and lossy-compression. The latter allows our approach to subsume the so-called geometrical compressibility as a special case. To the best knowledge of the authors, the established generalization bounds are the first of their kind for Information Bottleneck type encoders and representation learning. Finally, we partly exploit the theoretical results by introducing a new data-dependent prior. Numerical simulations illustrate the advantages of well-chosen such priors over classical priors used in IB.

https://papers.nips.cc/paper_files/paper/2023/file/054e9f9a286671ababa3213d6e59c1c2-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/21923-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/054e9f9a286671ababa3213d6e59c1c2-Supplemental-Conference.pdf

From Discrete Tokens to High-Fidelity Audio Using Multi-Band Diffusion

Main Conference Track

Robin San Roman, Yossi Adi, Antoine Deleforge, Romain Serizel, Gabriel Synnaeve, Alexandre Defossez

Deep generative models can generate high-fidelity audio conditioned on varioustypes of representations (e.g., mel-spectrograms, Mel-frequency Cepstral Coefficients(MFCC)). Recently, such models have been used to synthesize audiowaveforms conditioned on highly compressed representations. Although suchmethods produce impressive results, they are prone to generate audible artifactswhen the conditioning is flawed or imperfect. An alternative modeling approach isto use diffusion models. However, these have mainly been used as speech vocoders(i.e., conditioned on mel-spectrograms) or generating relatively low samplingrate signals. In this work, we propose a high-fidelity multi-band diffusion-basedframework that generates any type of audio modality (e.g., speech, music, environmentalsounds) from low-bitrate discrete representations. At equal bit rate,the proposed approach outperforms state-of-the-art generative techniques in termsof perceptual quality. Training and evaluation code are available on the facebookresearch/audiocraft github project. Samples are available on the followinglink (https://ai.honu.io/papers/mbd/).

https://papers.nips.cc/paper_files/paper/2023/file/054f771d614df12fe8def8ecdbe4e8e1-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/21274-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/054f771d614df12fe8def8ecdbe4e8e1-Supplemental-Conference.pdf

Fixing the NTK: From Neural Network Linearizations to Exact Convex Programs

Main Conference Track

Rajat Vadiraj Dwaraknath, Tolga Ergen, Mert Pilanci

Recently, theoretical analyses of deep neural networks have broadly focused on two directions: 1) Providing insight into neural network training by SGD in the limit of infinite hidden-layer width and infinitesimally small learning rate (also known as gradient flow) via the Neural Tangent Kernel (NTK), and 2) Globally optimizing the regularized training objective via cone-constrained convex reformulations of ReLU networks. The latter research direction also yielded an alternative formulation of the ReLU network, called a gated ReLU network, that is globally optimizable via efficient unconstrained convex programs. In this work, we interpret the convex program for this gated ReLU network as a Multiple Kernel Learning (MKL) model with a weighted data masking feature map and establish a connection to the NTK. Specifically, we show that for a particular choice of mask weights that do not depend on the learning targets, this kernel is equivalent to the NTK of the gated ReLU network on the training data. A consequence of this lack of dependence on the targets is that the NTK cannot perform better than the optimal MKL kernel on the training set. By using iterative reweighting, we improve the weights induced by the NTK to obtain the optimal MKL kernel which is equivalent to the solution of the exact convex reformulation of the gated ReLU network. We also provide several numerical simulations corroborating our theory. Additionally, we provide an analysis of the prediction error of the resulting optimal kernel via consistency results for the group lasso.

https://papers.nips.cc/paper_files/paper/2023/file/055fc19a3ce780b96cff15ffe738c1f1-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/21678-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/055fc19a3ce780b96cff15ffe738c1f1-Supplemental-Conference.zip

Birth of a Transformer: A Memory Viewpoint

Main Conference Track

Alberto Bietti, Vivien Cabannes, Diane Bouchacourt, Herve Jegou, Leon Bottou

Large language models based on transformers have achieved great empirical successes. However, as they are deployed more widely, there is a growing need to better understand their internal mechanisms in order to make them more reliable. These models appear to store vast amounts of knowledge from their training data, and to adapt quickly to new information provided in their context or prompt. We study how transformers balance these two types of knowledge by considering a synthetic setup where tokens are generated from either global or context-specific bigram distributions. By a careful empirical analysis of the training process on a simplified two-layer transformer, we illustrate the fast learning of global bigrams and the slower development of an "induction head" mechanism for the in-context bigrams. We highlight the role of weight matrices as associative memories, provide theoretical insights on how gradients enable their learning during training, and study the role of data-distributional properties.

https://papers.nips.cc/paper_files/paper/2023/file/0561738a239a995c8cd2ef0e50cfa4fd-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/19662-/bibtex

A Variational Perspective on High-Resolution ODEs

Main Conference Track

Hoomaan Maskan, Konstantinos Zygalakis, Alp Yurtsever

We consider unconstrained minimization of smooth convex functions. We propose a novel variational perspective using forced Euler-Lagrange equation that allows for studying high-resolution ODEs. Through this, we obtain a faster convergence rate for gradient norm minimization using Nesterov's accelerated gradient method. Additionally, we show that Nesterov's method can be interpreted as a rate-matching discretization of an appropriately chosen high-resolution ODE. Finally, using the results from the new variational perspective, we propose a stochastic method for noisy gradients. Several numerical experiments compare and illustrate our stochastic algorithm with state of the art methods.

https://papers.nips.cc/paper_files/paper/2023/file/0569458210c88d8db2985799da830d27-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/20103-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/0569458210c88d8db2985799da830d27-Supplemental-Conference.zip

What You See is What You Read? Improving Text-Image Alignment Evaluation

Main Conference Track

Michal Yarom, Yonatan Bitton, Soravit Changpinyo, Roee Aharoni, Jonathan Herzig, Oran Lang, Eran Ofek, Idan Szpektor

Automatically determining whether a text and a corresponding image are semantically aligned is a significant challenge for vision-language models, with applications in generative text-to-image and image-to-text tasks. In this work, we study methods for automatic text-image alignment evaluation. We first introduce SeeTRUE: a comprehensive evaluation set, spanning multiple datasets from both text-to-image and image-to-text generation tasks, with human judgements for whether a given text-image pair is semantically aligned. We then describe two automatic methods to determine alignment: the first involving a pipeline based on question generation and visual question answering models, and the second employing an end-to-end classification approach by finetuning multimodal pretrained models. Both methods surpass prior approaches in various text-image alignment tasks, with significant improvements in challenging cases that involve complex composition or unnatural images. Finally, we demonstrate how our approaches can localize specific misalignments between an image and a given text, and how they can be used to automatically re-rank candidates in text-to-image generation.

https://papers.nips.cc/paper_files/paper/2023/file/056e8e9c8ca9929cb6cf198952bf1dbb-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/22359-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/056e8e9c8ca9929cb6cf198952bf1dbb-Supplemental-Conference.zip

On the Robustness of Mechanism Design under Total Variation Distance

Main Conference Track

Anuran Makur, Marios Mertzanidis, Alexandros Psomas, Athina Terzoglou

We study the problem of designing mechanisms when agents' valuation functions are drawn from unknown and correlated prior distributions. In particular, we are given a prior distribution $D$, and we are interested in designing a (truthful) mechanism that has good performance for all "true distributions" that are close to $D$ in Total Variation (TV) distance. We show that DSIC and BIC mechanisms in this setting are strongly robust with respect to TV distance, for any bounded objective function $\mathcal{O}$, extending a recent result of Brustle et al. ([BCD20], EC 2020). At the heart of our result is a fundamental duality property of total variation distance. As direct applications of our result, we (i) demonstrate how to find approximately revenue-optimal and approximately BIC mechanisms for weakly dependent prior distributions; (ii) show how to find correlation-robust mechanisms when only ``noisy'' versions of marginals are accessible, extending recent results of Bei et. al. ([BGLT19], SODA 2019); (iii) prove that prophet-inequality type guarantees are preserved for correlated priors, recovering a variant of a result of D{\"u}tting and Kesselheim ([DK19], EC 2019) as a special case; (iv) give a new necessary condition for a correlated distribution to witness an infinite separation in revenue between simple and optimal mechanisms, complementing recent results of Psomas et al. ([PSCW22], NeurIPS 2022); (v) give a new condition for simple mechanisms to approximate revenue-optimal mechanisms for the case of a single agent whose type is drawn from a correlated distribution that can be captured by a Markov Random Field, complementing recent results of Cai and Oikonomou ([CO21], EC 2021).

https://papers.nips.cc/paper_files/paper/2023/file/058983528186511a74968e88a6d0ad63-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/20388-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/058983528186511a74968e88a6d0ad63-Supplemental-Conference.pdf

$\mathcal{M}^4$: A Unified XAI Benchmark for Faithfulness Evaluation of Feature Attribution Methods across Metrics, Modalities and Models

Datasets and Benchmarks Track

Xuhong Li, Mengnan Du, Jiamin Chen, Yekun Chai, Himabindu Lakkaraju, Haoyi Xiong

While Explainable Artificial Intelligence (XAI) techniques have been widely studied to explain predictions made by deep neural networks, the way to evaluate the faithfulness of explanation results remains challenging, due to the heterogeneity of explanations for various models and the lack of ground-truth explanations. This paper introduces an XAI benchmark named $\mathcal{M}^4$, which allows evaluating various input feature attribution methods using the same set of faithfulness metrics across multiple data modalities (images and texts) and network structures (ResNets, MobileNets, Transformers). A taxonomy for the metrics has been proposed as well. We first categorize commonly used XAI evaluation metrics into three groups based on the ground truth they require. We then implement classic and state-of-the-art feature attribution methods using InterpretDL and conduct extensive experiments to compare methods and gain insights. Extensive experiments have been conducted to provide holistic evaluations as benchmark baselines. Several interesting observations are noticed for designing attribution algorithms. The implementation of state-of-the-art explanation methods and evaluation metrics of $\mathcal{M}^4$ is publicly available at \url{https://github.com/PaddlePaddle/InterpretDL}.

https://papers.nips.cc/paper_files/paper/2023/file/05957c194f4c77ac9d91e1374d2def6b-Paper-Datasets_and_Benchmarks.pdf

https://papers.nips.cc/paper_files/paper/22790-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/05957c194f4c77ac9d91e1374d2def6b-Supplemental-Datasets_and_Benchmarks.zip

A generative model of the hippocampal formation trained with theta driven local learning rules

Main Conference Track

Tom M George, Kimberly L. Stachenfeld, Caswell Barry, Claudia Clopath, Tomoki Fukai

Advances in generative models have recently revolutionised machine learning. Meanwhile, in neuroscience, generative models have long been thought fundamental to animal intelligence. Understanding the biological mechanisms that support these processes promises to shed light on the relationship between biological and artificial intelligence. In animals, the hippocampal formation is thought to learn and use a generative model to support its role in spatial and non-spatial memory. Here we introduce a biologically plausible model of the hippocampal formation tantamount to a Helmholtz machine that we apply to a temporal stream of inputs. A novel component of our model is that fast theta-band oscillations (5-10 Hz) gate the direction of information flow throughout the network, training it akin to a high-frequency wake-sleep algorithm. Our model accurately infers the latent state of high-dimensional sensory environments and generates realistic sensory predictions. Furthermore, it can learn to path integrate by developing a ring attractor connectivity structure matching previous theoretical proposals and flexibly transfer this structure between environments. Whereas many models trade-off biological plausibility with generality, our model captures a variety of hippocampal cognitive functions under one biologically plausible local learning rule.

https://papers.nips.cc/paper_files/paper/2023/file/05ab457c7b769f01c2973e2a5ab66ad9-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/22627-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/05ab457c7b769f01c2973e2a5ab66ad9-Supplemental-Conference.pdf

Risk-Averse Model Uncertainty for Distributionally Robust Safe Reinforcement Learning

Main Conference Track

James Queeney, Mouhacine Benosman

Many real-world domains require safe decision making in uncertain environments. In this work, we introduce a deep reinforcement learning framework for approaching this important problem. We consider a distribution over transition models, and apply a risk-averse perspective towards model uncertainty through the use of coherent distortion risk measures. We provide robustness guarantees for this framework by showing it is equivalent to a specific class of distributionally robust safe reinforcement learning problems. Unlike existing approaches to robustness in deep reinforcement learning, however, our formulation does not involve minimax optimization. This leads to an efficient, model-free implementation of our approach that only requires standard data collection from a single training environment. In experiments on continuous control tasks with safety constraints, we demonstrate that our framework produces robust performance and safety at deployment time across a range of perturbed test environments.

https://papers.nips.cc/paper_files/paper/2023/file/05b63fa06784b71aab3939004e0f0a0d-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/19925-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/05b63fa06784b71aab3939004e0f0a0d-Supplemental-Conference.pdf

Optimal approximation using complex-valued neural networks

Main Conference Track

Paul Geuchen, Felix Voigtlaender

Complex-valued neural networks (CVNNs) have recently shown promising empirical success, for instance for increasing the stability of recurrent neural networks and for improving the performance in tasks with complex-valued inputs, such as MRI fingerprinting. While the overwhelming success of Deep Learning in the real-valued case is supported by a growing mathematical foundation, such a foundation is still largely lacking in the complex-valued case. We thus analyze the expressivity of CVNNs by studying their approximation properties. Our results yield the first quantitative approximation bounds for CVNNs that apply to a wide class of activation functions including the popular modReLU and complex cardioid activation functions. Precisely, our results apply to any activation function that is smooth but not polyharmonic on some non-empty open set; this is the natural generalization of the class of smooth and non-polynomial activation functions to the complex setting. Our main result shows that the approximation error scales as $m^{-k/(2n)}$ for $m \to \infty$ where $m$ is the number of neurons, $k$ the smoothness of the target function and $n$ is the (complex) input dimension. Under a natural continuity assumption, we show that this rate is optimal; we further discuss the optimality when dropping this assumption. Moreover, we prove that the problem of approximating $C^k$-functions using continuous approximation methods unavoidably suffers from the curse of dimensionality.

https://papers.nips.cc/paper_files/paper/2023/file/05b69cc4c8ff6e24c5de1ecd27223d37-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/22175-/bibtex

BayesDAG: Gradient-Based Posterior Inference for Causal Discovery

Main Conference Track

Yashas Annadani, Nick Pawlowski, Joel Jennings, Stefan Bauer, Cheng Zhang, Wenbo Gong

Bayesian causal discovery aims to infer the posterior distribution over causal models from observed data, quantifying epistemic uncertainty and benefiting downstream tasks. However, computational challenges arise due to joint inference over combinatorial space of Directed Acyclic Graphs (DAGs) and nonlinear functions. Despite recent progress towards efficient posterior inference over DAGs, existing methods are either limited to variational inference on node permutation matrices for linear causal models, leading to compromised inference accuracy, or continuous relaxation of adjacency matrices constrained by a DAG regularizer, which cannot ensure resulting graphs are DAGs. In this work, we introduce a scalable Bayesian causal discovery framework based on a combination of stochastic gradient Markov Chain Monte Carlo (SG-MCMC) and Variational Inference (VI) that overcomes these limitations. Our approach directly samples DAGs from the posterior without requiring any DAG regularization, simultaneously draws function parameter samples and is applicable to both linear and nonlinear causal models. To enable our approach, we derive a novel equivalence to the permutation-based DAG learning, which opens up possibilities of using any relaxed gradient estimator defined over permutations. To our knowledge, this is the first framework applying gradient-based MCMC sampling for causal discovery. Empirical evaluation on synthetic and real-world datasets demonstrate our approach's effectiveness compared to state-of-the-art baselines.

https://papers.nips.cc/paper_files/paper/2023/file/05cf28e3d3c9a179d789c55270fe6f72-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/19548-/bibtex

Bounce: Reliable High-Dimensional Bayesian Optimization for Combinatorial and Mixed Spaces

Main Conference Track

Leonard Papenmeier, Luigi Nardi, Matthias Poloczek

Impactful applications such as materials discovery, hardware design, neural architecture search, or portfolio optimization require optimizing high-dimensional black-box functions with mixed and combinatorial input spaces.While Bayesian optimization has recently made significant progress in solving such problems, an in-depth analysis reveals that the current state-of-the-art methods are not reliable. Their performances degrade substantially when the unknown optima of the function do not have a certain structure. To fill the need for a reliable algorithm for combinatorial and mixed spaces, this paper proposes Bounce that relies on a novel map of various variable types into nested embeddings of increasing dimensionality.Comprehensive experiments show that Bounce reliably achieves and often even improves upon state-of-the-art performance on a variety of high-dimensional problems.

https://papers.nips.cc/paper_files/paper/2023/file/05d2175de7ee637588d1b5ced8b15b32-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/20766-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/05d2175de7ee637588d1b5ced8b15b32-Supplemental-Conference.zip

Uniform-in-Time Wasserstein Stability Bounds for (Noisy) Stochastic Gradient Descent

Main Conference Track

Lingjiong Zhu, Mert Gurbuzbalaban, Anant Raj, Umut Simsekli

Algorithmic stability is an important notion that has proven powerful for deriving generalization bounds for practical algorithms. The last decade has witnessed an increasing number of stability bounds for different algorithms applied on different classes of loss functions. While these bounds have illuminated various properties of optimization algorithms, the analysis of each case typically required a different proof technique with significantly different mathematical tools. In this study, we make a novel connection between learning theory and applied probability and introduce a unified guideline for proving Wasserstein stability bounds for stochastic optimization algorithms. We illustrate our approach on stochastic gradient descent (SGD) and we obtain time-uniform stability bounds (i.e., the bound does not increase with the number of iterations) for strongly convex losses and non-convex losses with additive noise, where we recover similar results to the prior art or extend them to more general cases by using a single proof technique. Our approach is flexible and can be generalizable to other popular optimizers, as it mainly requires developing Lyapunov functions, which are often readily available in the literature. It also illustrates that ergodicity is an important component for obtaining time-uniform bounds -- which might not be achieved for convex or non-convex losses unless additional noise is injected to the iterates. Finally, we slightly stretch our analysis technique and prove time-uniform bounds for SGD under convex and non-convex losses (without additional additive noise), which, to our knowledge, is novel.

https://papers.nips.cc/paper_files/paper/2023/file/05d6b5b6901fb57d2c287e1d3ce6d63c-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/21554-/bibtex

Towards Generic Semi-Supervised Framework for Volumetric Medical Image Segmentation

Main Conference Track

Haonan Wang, Xiaomeng Li

Volume-wise labeling in 3D medical images is a time-consuming task that requires expertise. As a result, there is growing interest in using semi-supervised learning (SSL) techniques to train models with limited labeled data. However, the challenges and practical applications extend beyond SSL to settings such as unsupervised domain adaptation (UDA) and semi-supervised domain generalization (SemiDG). This work aims to develop a generic SSL framework that can handle all three settings. We identify two main obstacles to achieving this goal in the existing SSL framework: 1) the weakness of capturing distribution-invariant features; and 2) the tendency for unlabeled data to be overwhelmed by labeled data, leading to over-fitting to the labeled data during training. To address these issues, we propose an Aggregating & Decoupling framework. The aggregating part consists of a Diffusion encoder that constructs a "common knowledge set" by extracting distribution-invariant features from aggregated information from multiple distributions/domains. The decoupling part consists of three decoders that decouple the training process with labeled and unlabeled data, thus avoiding over-fitting to labeled data, specific domains and classes. We evaluate our proposed framework on four benchmark datasets for SSL, Class-imbalanced SSL, UDA and SemiDG. The results showcase notable improvements compared to state-of-the-art methods across all four settings, indicating the potential of our framework to tackle more challenging SSL scenarios. Code and models are available at: https://github.com/xmed-lab/GenericSSL.

https://papers.nips.cc/paper_files/paper/2023/file/05dc08730e32441edff52b0fa6caab5f-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/19625-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/05dc08730e32441edff52b0fa6caab5f-Supplemental-Conference.pdf

Stochastic Distributed Optimization under Average Second-order Similarity: Algorithms and Analysis

Main Conference Track

Dachao Lin, Yuze Han, Haishan Ye, Zhihua Zhang

We study finite-sum distributed optimization problems involving a master node and $n-1$ local nodes under the popular $\delta$-similarity and $\mu$-strong convexity conditions. We propose two new algorithms, SVRS and AccSVRS, motivated by previous works. The non-accelerated SVRS method combines the techniques of gradient sliding and variance reduction and achieves a better communication complexity of $\tilde{\mathcal{O}}(n {+} \sqrt{n}\delta/\mu)$ compared to existing non-accelerated algorithms. Applying the framework proposed in Katyusha X, we also develop a directly accelerated version named AccSVRS with the $\tilde{\mathcal{O}}(n {+} n^{3/4}\sqrt{\delta/\mu})$ communication complexity. In contrast to existing results, our complexity bounds are entirely smoothness-free and exhibit superiority in ill-conditioned cases. Furthermore, we establish a nearly matched lower bound to verify the tightness of our AccSVRS method.

https://papers.nips.cc/paper_files/paper/2023/file/05e552739c2629f3324c1063a382b4bd-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/22682-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/05e552739c2629f3324c1063a382b4bd-Supplemental-Conference.pdf

PolyDiffuse: Polygonal Shape Reconstruction via Guided Set Diffusion Models

Main Conference Track

Jiacheng Chen, Ruizhi Deng, Yasutaka Furukawa

This paper presents \textit{PolyDiffuse}, a novel structured reconstruction algorithm that transforms visual sensor data into polygonal shapes with Diffusion Models (DM), an emerging machinery amid exploding generative AI, while formulating reconstruction as a generation process conditioned on sensor data. The task of structured reconstruction poses two fundamental challenges to DM: 1) A structured geometry is a ''set'' (e.g., a set of polygons for a floorplan geometry), where a sample of $N$ elements has $N!$ different but equivalent representations, making the denoising highly ambiguous; and 2) A ''reconstruction'' task has a single solution, where an initial noise needs to be chosen carefully, while any initial noise works for a generation task.Our technical contribution is the introduction of a Guided Set Diffusion Model where 1) the forward diffusion process learns \textit{guidance networks} to control noise injection so that one representation of a sample remains distinct from its other permutation variants, thus resolving denoising ambiguity; and 2) the reverse denoising process reconstructs polygonal shapes, initialized and directed by the guidance networks, as a conditional generation process subject to the sensor data.We have evaluated our approach for reconstructing two types of polygonal shapes: floorplan as a set of polygons and HD map for autonomous cars as a set of polylines.Through extensive experiments on standard benchmarks, we demonstrate that PolyDiffuse significantly advances the current state of the art and enables broader practical applications. The code and data are available on our project page: https://poly-diffuse.github.io.

https://papers.nips.cc/paper_files/paper/2023/file/05f0e2fa003602db2d98ca72b79dec51-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/21588-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/05f0e2fa003602db2d98ca72b79dec51-Supplemental-Conference.pdf

Can You Rely on Your Model Evaluation? Improving Model Evaluation with Synthetic Test Data

Main Conference Track

Boris van Breugel, Nabeel Seedat, Fergus Imrie, Mihaela van der Schaar

Evaluating the performance of machine learning models on diverse and underrepresented subgroups is essential for ensuring fairness and reliability in real-world applications. However, accurately assessing model performance becomes challenging due to two main issues: (1) a scarcity of test data, especially for small subgroups, and (2) possible distributional shifts in the model's deployment setting, which may not align with the available test data. In this work, we introduce 3S Testing, a deep generative modeling framework to facilitate model evaluation by generating synthetic test sets for small subgroups and simulating distributional shifts. Our experiments demonstrate that 3S-Testing outperforms traditional baselines---including real test data alone---in estimating model performance on minority subgroups and under plausible distributional shifts. In addition, 3S offers intervals around its performance estimates, exhibiting superior coverage of the ground truth compared to existing approaches. Overall, these results raise the question of whether we need a paradigm shift away from limited real test data towards synthetic test data.

https://papers.nips.cc/paper_files/paper/2023/file/05fb0f4e645cad23e0ab59d6b9901428-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/20383-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/05fb0f4e645cad23e0ab59d6b9901428-Supplemental-Conference.pdf

Rethinking the Backward Propagation for Adversarial Transferability

Main Conference Track

Wang Xiaosen, Kangheng Tong, Kun He

Transfer-based attacks generate adversarial examples on the surrogate model, which can mislead other black-box models without access, making it promising to attack real-world applications. Recently, several works have been proposed to boost adversarial transferability, in which the surrogate model is usually overlooked. In this work, we identify that non-linear layers (e.g., ReLU, max-pooling, etc.) truncate the gradient during backward propagation, making the gradient w.r.t. input image imprecise to the loss function. We hypothesize and empirically validate that such truncation undermines the transferability of adversarial examples. Based on these findings, we propose a novel method called Backward Propagation Attack (BPA) to increase the relevance between the gradient w.r.t. input image and loss function so as to generate adversarial examples with higher transferability. Specifically, BPA adopts a non-monotonic function as the derivative of ReLU and incorporates softmax with temperature to smooth the derivative of max-pooling, thereby mitigating the information loss during the backward propagation of gradients. Empirical results on the ImageNet dataset demonstrate that not only does our method substantially boost the adversarial transferability, but it is also general to existing transfer-based attacks. Code is available at https://github.com/Trustworthy-AI-Group/RPA.

https://papers.nips.cc/paper_files/paper/2023/file/05fe0c633ae41756540dba2a99a36306-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/20937-/bibtex

Bullying10K: A Large-Scale Neuromorphic Dataset towards Privacy-Preserving Bullying Recognition

Datasets and Benchmarks Track

Yiting Dong, Yang Li, Dongcheng Zhao, Guobin Shen, Yi Zeng

The prevalence of violence in daily life poses significant threats to individuals' physical and mental well-being. Using surveillance cameras in public spaces has proven effective in proactively deterring and preventing such incidents. However, concerns regarding privacy invasion have emerged due to their widespread deployment.To address the problem, we leverage Dynamic Vision Sensors (DVS) cameras to detect violent incidents and preserve privacy since it captures pixel brightness variations instead of static imagery. We introduce the Bullying10K dataset, encompassing various actions, complex movements, and occlusions from real-life scenarios. It provides three benchmarks for evaluating different tasks: action recognition, temporal action localization, and pose estimation. With 10,000 event segments, totaling 12 billion events and 255 GB of data, Bullying10K contributes significantly by balancing violence detection and personal privacy persevering. And it also poses a challenge to the neuromorphic dataset. It will serve as a valuable resource for training and developing privacy-protecting video systems. The Bullying10K opens new possibilities for innovative approaches in these domains.

https://papers.nips.cc/paper_files/paper/2023/file/05ffe69463062b7f9fb506c8351ffdd7-Paper-Datasets_and_Benchmarks.pdf

https://papers.nips.cc/paper_files/paper/22580-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/05ffe69463062b7f9fb506c8351ffdd7-Supplemental-Datasets_and_Benchmarks.pdf

Compression with Bayesian Implicit Neural Representations

Main Conference Track

Zongyu Guo, Gergely Flamich, Jiajun He, Zhibo Chen, José Miguel Hernández-Lobato

Many common types of data can be represented as functions that map coordinates to signal values, such as pixel locations to RGB values in the case of an image. Based on this view, data can be compressed by overfitting a compact neural network to its functional representation and then encoding the network weights. However, most current solutions for this are inefficient, as quantization to low-bit precision substantially degrades the reconstruction quality. To address this issue, we propose overfitting variational Bayesian neural networks to the data and compressing an approximate posterior weight sample using relative entropy coding instead of quantizing and entropy coding it. This strategy enables direct optimization of the rate-distortion performance by minimizing the $\beta$-ELBO, and target different rate-distortion trade-offs for a given network architecture by adjusting $\beta$. Moreover, we introduce an iterative algorithm for learning prior weight distributions and employ a progressive refinement process for the variational posterior that significantly enhances performance. Experiments show that our method achieves strong performance on image and audio compression while retaining simplicity.

https://papers.nips.cc/paper_files/paper/2023/file/060b2af0081a460f7f466f7f174d9052-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/21267-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/060b2af0081a460f7f466f7f174d9052-Supplemental-Conference.pdf

Towards Unbounded Machine Unlearning

Main Conference Track

Meghdad Kurmanji, Peter Triantafillou, Jamie Hayes, Eleni Triantafillou

Deep machine unlearning is the problem of 'removing' from a trained neural network a subset of its training set. This problem is very timely and has many applications, including the key tasks of removing biases (RB), resolving confusion (RC) (caused by mislabelled data in trained models), as well as allowing users to exercise their 'right to be forgotten' to protect User Privacy (UP). This paper is the first, to our knowledge, to study unlearning for different applications (RB, RC, UP), with the view that each has its own desiderata, definitions for 'forgetting' and associated metrics for forget quality. For UP, we propose a novel adaptation of a strong Membership Inference Attack for unlearning. We also propose SCRUB, a novel unlearning algorithm, which is the only method that is consistently a top performer for forget quality across the different application-dependent metrics for RB, RC, and UP. At the same time, SCRUB is also consistently a top performer on metrics that measure model utility (i.e. accuracy on retained data and generalization), and is more efficient than previous work. The above are substantiated through a comprehensive empirical evaluation against previous state-of-the-art.

https://papers.nips.cc/paper_files/paper/2023/file/062d711fb777322e2152435459e6e9d9-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/21511-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/062d711fb777322e2152435459e6e9d9-Supplemental-Conference.zip

Collaborative Learning via Prediction Consensus

Main Conference Track

Dongyang Fan, Celestine Mendler-Dünner, Martin Jaggi

We consider a collaborative learning setting where the goal of each agent is to improve their own model by leveraging the expertise of collaborators, in addition to their own training data. To facilitate the exchange of expertise among agents, we propose a distillation-based method leveraging shared unlabeled auxiliary data, which is pseudo-labeled by the collective. Central to our method is a trust weighting scheme that serves to adaptively weigh the influence of each collaborator on the pseudo-labels until a consensus on how to label the auxiliary data is reached. We demonstrate empirically that our collaboration scheme is able to significantly boost individual models’ performance in the target domain from which the auxiliary data is sampled. At the same time, it can provably mitigate the negative impact of bad models on the collective. By design, our method adeptly accommodates heterogeneity in model architectures and substantially reduces communication overhead compared to typical collaborative learning methods.

https://papers.nips.cc/paper_files/paper/2023/file/065e259a1d2d955e63b99aac6a3a3081-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/22332-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/065e259a1d2d955e63b99aac6a3a3081-Supplemental-Conference.pdf

Identification of Nonlinear Latent Hierarchical Models

Main Conference Track

Lingjing Kong, Biwei Huang, Feng Xie, Eric Xing, Yuejie Chi, Kun Zhang

Identifying latent variables and causal structures from observational data is essential to many real-world applications involving biological data, medical data, and unstructured data such as images and languages. However, this task can be highly challenging, especially when observed variables are generated by causally related latent variables and the relationships are nonlinear. In this work, we investigate the identification problem for nonlinear latent hierarchical causal models in which observed variables are generated by a set of causally related latent variables, and some latent variables may not have observed children. We show that the identifiability of causal structures and latent variables (up to invertible transformations) can be achieved under mild assumptions: on causal structures, we allow for multiple paths between any pair of variables in the graph, which relaxes latent tree assumptions in prior work; on structural functions, we permit general nonlinearity and multi-dimensional continuous variables, alleviating existing work's parametric assumptions. Specifically, we first develop an identification criterion in the form of novel identifiability guarantees for an elementary latent variable model. Leveraging this criterion, we show that both causal structures and latent variables of the hierarchical model can be identified asymptotically by explicitly constructing an estimation procedure. To the best of our knowledge, our work is the first to establish identifiability guarantees for both causal structures and latent variables in nonlinear latent hierarchical models.

https://papers.nips.cc/paper_files/paper/2023/file/065ef23a944b3995de7dd4a3e203d133-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/20693-/bibtex

Sample Efficient Reinforcement Learning in Mixed Systems through Augmented Samples and Its Applications to Queueing Networks

Main Conference Track

Honghao Wei, Xin Liu, Weina Wang, Lei Ying

This paper considers a class of reinforcement learning problems, which involve systems with two types of states: stochastic and pseudo-stochastic. In such systems, stochastic states follow a stochastic transition kernel while the transitions of pseudo-stochastic states are deterministic {\em given} the stochastic states/transitions. We refer to such systems as mixed systems, which are widely used in various applications, including Manufacturing systems, communication networks, and queueing networks. We propose a sample-efficient RL method that accelerates learning by generating augmented data samples. The proposed algorithm is data-driven (model-free), but it learns the policy from data samples from both real and augmented samples. This method significantly improves learning by reducing the sample complexity such that the dataset only needs to have sufficient coverage of the stochastic states. We analyze the sample complexity of the proposed method under Fitted Q Iteration (FQI) and demonstrate that the optimality gap decreases as $O\left(\sqrt{\frac{1}{n}}+\sqrt{\frac{1}{m}}\right),$ where $n$ represents the number of real samples, and $m$ is the number of augmented samples per real sample. It is important to note that without augmented samples, the optimality gap is $O(1)$ due to the insufficient data coverage of the pseudo-stochastic states. Our experimental results on multiple queueing network applications confirm that the proposed method indeed significantly accelerates both deep Q-learning and deep policy gradient.

https://papers.nips.cc/paper_files/paper/2023/file/0663a39baab211328fc865f91abc75ab-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/22719-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/0663a39baab211328fc865f91abc75ab-Supplemental-Conference.zip

Temporal Graph Benchmark for Machine Learning on Temporal Graphs

Datasets and Benchmarks Track

Shenyang Huang, Farimah Poursafaei, Jacob Danovitch, Matthias Fey, Weihua Hu, Emanuele Rossi, Jure Leskovec, Michael Bronstein, Guillaume Rabusseau, Reihaneh Rabbany

We present the Temporal Graph Benchmark (TGB), a collection of challenging and diverse benchmark datasets for realistic, reproducible, and robust evaluation of machine learning models on temporal graphs. TGB datasets are of large scale, spanning years in duration, incorporate both node and edge-level prediction tasks and cover a diverse set of domains including social, trade, transaction, and transportation networks. For both tasks, we design evaluation protocols based on realistic use-cases. We extensively benchmark each dataset and find that the performance of common models can vary drastically across datasets. In addition, on dynamic node property prediction tasks, we show that simple methods often achieve superior performance compared to existing temporal graph models. We believe that these findings open up opportunities for future research on temporal graphs. Finally, TGB provides an automated machine learning pipeline for reproducible and accessible temporal graph research, including data loading, experiment setup and performance evaluation. TGB will be maintained and updated on a regular basis and welcomes community feedback. TGB datasets, data loaders, example codes, evaluation setup, and leaderboards are publicly available at https://tgb.complexdatalab.com/.

https://papers.nips.cc/paper_files/paper/2023/file/066b98e63313162f6562b35962671288-Paper-Datasets_and_Benchmarks.pdf

https://papers.nips.cc/paper_files/paper/22446-/bibtex

Navigating Data Heterogeneity in Federated Learning: A Semi-Supervised Federated Object Detection

Main Conference Track

Taehyeon Kim, Eric Lin, Junu Lee, Christian Lau, Vaikkunth Mugunthan

Federated Learning (FL) has emerged as a potent framework for training models across distributed data sources while maintaining data privacy. Nevertheless, it faces challenges with limited high-quality labels and non-IID client data, particularly in applications like autonomous driving. To address these hurdles, we navigate the uncharted waters of Semi-Supervised Federated Object Detection (SSFOD). We present a pioneering SSFOD framework, designed for scenarios where labeled data reside only at the server while clients possess unlabeled data. Notably, our method represents the inaugural implementation of SSFOD for clients with 0% labeled non-IID data, a stark contrast to previous studies that maintain some subset of labels at each client. We propose FedSTO, a two-stage strategy encompassing Selective Training followed by Orthogonally enhanced full-parameter training, to effectively address data shift (e.g. weather conditions) between server and clients. Our contributions include selectively refining the backbone of the detector to avert overfitting, orthogonality regularization to boost representation divergence, and local EMA-driven pseudo label assignment to yield high-quality pseudo labels. Extensive validation on prominent autonomous driving datasets (BDD100K, Cityscapes, and SODA10M) attests to the efficacy of our approach, demonstrating state-of-the-art results. Remarkably, FedSTO, using just 20-30% of labels, performs nearly as well as fully-supervised centralized training methods.

https://papers.nips.cc/paper_files/paper/2023/file/066e4dbfeccb5dc2851acd5eca584937-Paper-Conference.pdf

https://papers.nips.cc/paper_files/paper/22736-/bibtex

https://papers.nips.cc/paper_files/paper/2023/file/066e4dbfeccb5dc2851acd5eca584937-Supplemental-Conference.pdf