Samy Bengio

Reasoning’s Razor: Reasoning Improves Accuracy but Hurts Recall at Critical Operating Points in Safety and Hallucination Detection EACL 2026 GSM-Symbolic: Understanding the Limitations of Mathematical Reasoning in Large Language Models ICLR 2025 TiC-LM: A Web-Scale Benchmark for Time-Continual LLM Pretraining ACL 2025 Generalization on the Unseen, Logic Reasoning and Degree Curriculum JMLR 2024 How Far Can Transformers Reason? The Globality Barrier and Inductive Scratchpad NeurIPS 2024 What Algorithms can Transformers Learn? A Study in Length Generalization ICLR 2024 When can transformers reason with abstract symbols? ICLR 2024 Continuous pseudo-labeling from the start ICLR 2023 Generalization on the Unseen, Logic Reasoning and Degree Curriculum ICML 2023 Transformers learn through gradual rank increase NeurIPS 2023 Are All Layers Created Equal? JMLR 2022 Learning to Reason with Neural Networks: Generalization, Unseen Data and Boolean Measures NeurIPS 2022 Improving Anytime Prediction with Parallel Cascaded Networks and a Temporal-Difference Loss NeurIPS 2021 Learnable Fourier Features for Multi-dimensional Spatial Positional Encoding NeurIPS 2021 Fantastic Generalization Measures and Where to Find Them ICLR 2020 Identity Crisis: Memorization and Generalization Under Extreme Overparameterization ICLR 2020 Memory Based Trajectory-conditioned Policies for Learning from Sparse Rewards NeurIPS 2020 Rapid Learning or Feature Reuse? Towards Understanding the Effectiveness of MAML ICLR 2020 Area Attention ICML 2019 Predicting the Generalization Gap in Deep Networks with Margin Distributions ICLR 2019 Transfusion: Understanding Transfer Learning for Medical Imaging NeurIPS 2019 You Look Twice: GaterNet for Dynamic Filter Selection in CNNs CVPR 2019 Content preserving text generation with attribute controls NeurIPS 2018 Fast Decoding in Sequence Models Using Discrete Latent Variables ICML 2018 Insights on representational similarity in neural networks with canonical correlation NeurIPS 2018 Large Margin Deep Networks for Classification NeurIPS 2018 Context-Aware Captions From Context-Agnostic Supervision CVPR 2017 Device Placement Optimization with Reinforcement Learning ICML 2017 Sharp Minima Can Generalize For Deep Nets ICML 2017 Tacotron: Towards End-to-End Speech Synthesis Interspeech 2017 ADIOS: Architectures Deep In Output Space ICML 2016 An Online Sequence-to-Sequence Model Using Partial Conditioning NeurIPS 2016 Can Active Memory Replace Attention? NeurIPS 2016 Generating Sentences from a Continuous Space CoNLL 2016 LLORMA: Local Low-Rank Matrix Approximation JMLR 2016 Reward Augmented Maximum Likelihood for Neural Structured Prediction NeurIPS 2016 Learning Semantic Relationships for Better Action Retrieval in Images CVPR 2015 Scheduled Sampling for Sequence Prediction with Recurrent Neural Networks NeurIPS 2015 Show and Tell: A Neural Image Caption Generator CVPR 2015 Training Highly Multiclass Classifiers JMLR 2014 DeViSE: A Deep Visual-Semantic Embedding Model NeurIPS 2013 Label Embedding Trees for Large Multi-Class Tasks NeurIPS 2010 Large Scale Online Learning of Image Similarity Through Ranking JMLR 2010 Why Does Unsupervised Pre-training Help Deep Learning? JMLR 2010 An Online Algorithm for Large Scale Image Similarity Learning NeurIPS 2009 Group Sparse Coding NeurIPS 2009 The Need for Open Source Software in Machine Learning JMLR 2007 Investigating Lexical Substitution Scoring for Subtitle Generation CoNLL 2006 The Difficulty of Training Deep Architectures and the Effect of Unsupervised Pre-Training AISTATS 2005 SVMTorch: Support Vector Machines for Large-Scale Regression Problems JMLR 2001