[06/11/25] Excited to announce the release of our frontier video understanding model, VJEPA-2 ! Checkout our model code , weights and cool demos ! [06/11/25] Happy to release the Physical World Reasoning leaderboard, along with a new dataset, MVPBench (Paper ) for assessing video understanding capabilities of modern VLMs. [01/06/24] Serving as a Senior Area Chair at ACL 2024. [16/05/24] Excited to announce our large multimodal language model, Chameleon , is now released on arxiv ! We have also released the code and model weights , as per our Open Source Strategy ! [10/18/23] MLRC 2023 challenge goes live in partnership with TMLR! Read our blog post for more information. [07/10/23] Excited to share that our paper “Language model acceptability judgements are not always robust to context” has received Outstanding Paper Award [06/01/23] New paper: “Language model acceptability judgements are not always robust to context” , which is now accepted as a long paper in ACL 2023, Toronto. [11/02/22] Successfully defended my PhD! Checkout my thesis here . [08/29/22] Excited to announce a major life event: I’m starting today as a Research Scientist (Speech & NLP) in Meta AI New York! [08/19/22] Happy to announce yet another Machine Learning Reproducibility Challenge, the MLRC 2022 ! This is our six edition! [01/10/21] Happy to update that our paper “Masked Language Modeling and the Distributional Hypothesis: Order Word Matters Pre-training for Little” is accepted as a long paper at EMNLP 2021! [01/09/21] Happy to announce the new iteration of ML Reproducibility Challenge 2021 , which has now enlarged to cover 9 top ML conferences! Submit your reports through Feb 2022! [03/07/21] On a personal news, got married to my sweetheart Atrayee this July! [02/07/21] Thrilled to share that our paper UnNatural Language Inference has received Outstanding Paper Award at ACL 2021! Deeply honored! [15/04/21] Announcing the pre-print of our paper “Masked Language Modeling and the Distributional Hypothesis: Order Word Matters Pre-training for Little” . We find RoBERTa trained with sentence word order shuffled data performs remarkably close to natural word order pre-trained models on several downstream and probing tasks! [01/06/21] Excited to announce that our paper “UnNatural Language Inference” , has been accepted to ACL 2021 (Long paper, Oral), where we stumble upon the weird language understanding mechanisms employed by NLU models! [02/10/20] Happy to announce our paper “Measuring Systematic Generalization in Neural Proof Generation with Transformers” is accepted at NeurIPS 2020! [05/09/20] Excited to announce the 2020 edition of the ML Reproducibility Challenge ! We now cover 7 major ML conferences, do check it out! [05/08/20] We released a new blog post on ML Reproducibility Tools and Best Practices . Check it out! [30/04/20] Public release of our new multi-task graph dataset, GraphLogblog post for more information. [08/04/20] Report on NeurIPS 2019 Reproducibility Program published on arxiv. We have also published our thoughts on Designing the Reproducibility Program for NeurIPS 2020 on Medium. [15/04/20] Excited to announce two papers accepted to ACL 2020! Probing Linguistic Systematicity and Learning an unreferenced metric for online Dialog evaluation . [01/12/19] Co-organizing NeurIPS 2019 ML Retrospectives Workshop [01/09/19] Co-organizing NeurIPS 2019 Reproducibility Challenge and honored to be the NeurIPS 2019 Reproducibility Co-Chair. [28/01/19] Excited to join Facebook AI Research (FAIR) as PhD Intern! [14/08/19] Our paper CLUTRR: A Diagnostic Benchmark for Inductive Reasoning from Text [28/09/18] Co-organizing ICLR Reproducibility Challenge , 2019 [04/09/18] Starting PhD at McGill University , advised by Dr Joelle Pineau and Dr William L. Hamilton , from Fall 2018. [31/08/18] Our paper on A Hierarchical Neural Attention-based Text Classifier accepted at EMNLP 2018! [01/06/18] Intern-ing at Samsung Advanced Institute of Technology for the Summer! [01/02/18] Our paper on Ethics in Data Driven Dialog Systems accepted at AAAI/ACM conference on Ethics & Safety. XMin theme and Hugo.