Structural Pruning via Latency-Saliency Knapsack
Comments Accepted by NeurIPS 2022. arXiv admin note: substantial text overlap with arXiv:2110.10811
期刊&会议
NeurIPS
Conference on Neural Information Processing Systems · 会议 · Machine Learning
Unsupervised Multi-View Object Segmentation Using Radiance Field Propagation
Comments 23 pages, 14 figures, NeurIPS 2022
Where Should I Spend My FLOPS? Efficiency Evaluations of Visual Pre-training Methods
Comments 11 pages. 36th Conference on Neural Information Processing Systems, Workshop on Self-Supervised Learning (2022)
Lethal Dose Conjecture on Data Poisoning
Comments 36th Conference on Neural Information Processing Systems (NeurIPS 2022)
BackdoorBench: A Comprehensive Benchmark of Backdoor Learning
Comments Accepted at NeurIPS 2022 Datasets and Benchmarks Track; 44 pages; 8 backdoor attacks; 9 backdoor defenses; 8,000 pairs of attack-defense evaluations; several analysis and 5 analysis tools
Identifiability of deep generative models without auxiliary information
Comments 34 pages, 9 figures, to appear in NeurIPS 2022
A Unified Convergence Theorem for Stochastic Optimization Methods
Comments Accepted in the 36th Conference on Neural Information Processing Systems (NeurIPS 2022)
Bring Your Own Algorithm for Optimal Differentially Private Stochastic Minimax Optimization
Comments NeurIPS 2022
Neural Basis Models for Interpretability
Comments 17 pages including appendix. v2 includes link to source code available at https://github.com/facebookresearch/nbm-spam. v3 includes updates to baseline results, v4 updated for NeurIPS camera ready
Scalable Interpretability via Polynomials
Comments 26 pages including appendix. v2 includes source code link at https://github.com/facebookresearch/nbm-spam, v3 fixes to baseline results in Table 1, v4 update for NeurIPS camera ready
Hybrid Neural Autoencoders for Stimulus Encoding in Visual and Other Sensory Neuroprostheses
Comments NeurIPS 2022 camera ready revision
Active Surrogate Estimators: An Active Learning Approach to Label-Efficient Model Evaluation
Comments Accepted for publication at NeurIPS 2022
How Would The Viewer Feel? Estimating Wellbeing From Video Scenarios
Comments NeurIPS 2022; datasets available at https://github.com/hendrycks/emodiversity/
SQ Lower Bounds for Learning Single Neurons with Massart Noise
Comments To appear in NeurIPS 2022
HUMANISE: Language-conditioned Human Motion Generation in 3D Scenes
Comments Accepted by NeurIPS 2022
Hierarchical Normalization for Robust Monocular Depth Estimation
Comments Accepted to NeurIPS 2022
Planning for Sample Efficient Imitation Learning
Comments NeurIPS 2022
Robust Imitation of a Few Demonstrations with a Backwards Model
Comments Conference on Neural Information Processing Systems (NeurIPS) 2022
Self-explaining deep models with logic rule reasoning
Comments 26 pages including reference, checklist, and appendix. Accepted in NeurIPS 2022
Anticipating Performativity by Predicting from Predictions
Comments to appear at NeurIPS 2022, revision corresponds to camera ready version
Robust Reinforcement Learning using Offline Data
Comments Appeared in Neural Information Processing Systems (NeurIPS) 2022
Adversarial Robustness is at Odds with Lazy Training
Comments NeurIPS 2022
When Does Group Invariant Learning Survive Spurious Correlations?
Comments 25 pages, 3 figures. Accepted by NeurIPS 2022
Bootstrapped Transformer for Offline Reinforcement Learning
Comments Accepted in NeurIPS 2022
Anchor-Changing Regularized Natural Policy Gradient for Multi-Objective Reinforcement Learning
Comments 36th Conference on Neural Information Processing Systems (NeurIPS 2022)
GlanceNets: Interpretabile, Leak-proof Concept-based Models
Comments 36th Conference on Neural Information Processing Systems (NeurIPS 2022)
3DILG: Irregular Latent Grids for 3D Generative Modeling
Comments Accepted at NeurIPS 2022. Project page: https://1zb.github.io/3DILG
Low-rank lottery tickets: finding efficient low-rank neural networks via matrix differential equations
Journal ref Proceedings NeurIPS 2022
Recovering Private Text in Federated Learning of Language Models
Comments NeurIPS 2022. Code is publicly available at https://github.com/Princeton-SysML/FILM. v2 added discussion and evaluation of defenses
Unsupervised Cross-Task Generalization via Retrieval Augmentation
Comments Accepted to NeurIPS 2022. Website: https://inklab.usc.edu/ReCross/
