Prépublications

  • The Markovian Thinker
    Milad Aghajohari*, , Amirhossein Kazemnejad*, , Alessandro Sordoni, Aaron Courville et Siva Reddy
    In ArXiv, 2025.
    #NLP, #RL
    [arXiv]

  • Just-in-time Episodic Feedback Hinter: Leveraging Offline Knowledge to Improve LLM Agents Adaptation
    , Aman Jaiswal, Patrice Bechard, Oleh Shliazhko, Orlando Marquez Ayala, , Massimo Caccia, Alexandre Drouin, et Alexandre Lacoste
    In ArXiv, 2025.
    #NLP, #RL
    [arXiv]

  • GRPO-λ: Credit Assignment improves LLM Reasoning
    Prasanna Parthasarathi*, , Boxing Chen, Yufei Cui et
    In ArXiv, 2025.
    #RL, #NLP
    [arXiv]

  • NovoMolGen: Rethinking Molecular Language Model Pretraining
    , , Quentin Fournier, Nirav Pravinbhai Bhatt et
    In ArXiv, 2025.
    #NLP, #Other
    [arXiv]

  • CADmium: Fine-Tuning Code Language Models for Text-Driven Sequential CAD Design
    , , Jay Pathak, Quentin Fournier et
    In ArXiv, 2025.
    #NLP
    [arXiv]

  • Did I Faithfully Say What I Thought? Bridging the Gap Between Neural Activity and Self-Explanations in Large Language Models
    , Jean-Noel Vittaut, Nicolas Chesneau, et Marie-Jeanne Lesot
    In ArXiv, 2025.
    #NLP
    [arXiv]

  • Structure-Aligned Protein Language Model
    Can Chen, , Robert M. Vernon, Christopher James Langmead, Yoshua Bengio et Quentin Fournier
    In ArXiv, 2025.
    #NLP, #Other
    [arXiv]

  • Monitoring morphometric drift in lifelong learning segmentation of the spinal cord
    Enamundram Naga Karthik, Sandrine Bédard, Jan Valošek, Christoph S. Aigner, Elise Bannier, Josef Bednařík, Virginie Callot, Anna Combes, Armin Curt, Gergely David, Falk Eippert, Lynn Farner, Michael G Fehlings, Patrick Freund, Tobias Granberg, Cristina Granziera, RHSCIR Network Imaging Group, Ulrike Horn, Tomáš Horák, Suzanne Humphreys, Markus Hupp, Anne Kerbrat, Nawal Kinany, Shannon Kolind, Petr Kudlička, Anna Lebret, Lisa Eunyoung Lee, Caterina Mainero, Allan R. Martin, Megan McGrath, Govind Nair, Kristin P. O'Grady, Jiwon Oh, Russell Ouellette, Nikolai Pfender, Dario Pfyffer, Pierre-François Pradat, Alexandre Prat, Emanuele Pravatà, Daniel S. Reich, Ilaria Ricchi, Naama Rotem-Kohavi, Simon Schading-Sassenhausen, Maryam Seif, Andrew Smith, Seth A Smith, Grace Sweeney, Roger Tam, Anthony Traboulsee, Constantina Andrada Treaba, Charidimos Tsagkas, Zachary Vavasour, Dimitri Van De Ville, Kenneth Arnold Weber II, et Julien Cohen-Adad
    In ArXiv, 2025.
    #NLP
    [arXiv]

  • Too Big to Fool: Resisting Deception in Language Models
    , Mats Leon Richter, Juan Rodriguez, , et Maxime Gasse
    In ArXiv, 2024.
    #NLP
    [arXiv]

  • Interpretability Needs a New Paradigm
    , Himabindu Lakkaraju, Siva Reddy et
    In ArXiv, 2024.
    #NLP, #DL
    [arXiv]

Articles de conférence et de revue

2025

  1. How to Train Your LLM Web Agent: A Statistical Diagnosis
    Dheeraj Vattikonda, Santhoshi Ravichandran, Emiliano Penaloza, , , Thibault Le Sellier de Chezelles, Nicolas Gontier, Miguel Muñoz-Mármol, Sahar Omidi Shayegan, Stefania Raimondo, Xue Liu, Alexandre Drouin, Laurent Charlin, Alexandre Piché, Alexandre Lacoste et Massimo Caccia
    Conference on Neural Information Processing Systems (NeurIPS), 2025.
    #NLP, #RL
    [arXiv]

  2. Rendering-Aware Reinforcement Learning for Vector Graphics Generation
    Juan A. Rodriguez, Haotian Zhang, Abhay Puri, Aarash Feizi, Rishav Pramanik, Pascal Wichmann, Arnab Mondal, , Rabiul Awal, Perouz Taslakian, Spandana Gella, Sai Rajeswar, David Vazquez, Christopher Pal et Marco Pedersoli
    Conference on Neural Information Processing Systems (NeurIPS), 2025.
    #NLP, #RL
    [arXiv]

  3. Steering Large Language Model Activations in Sparse Spaces
    Reza Bayat*, , Mohammad Pezeshki, et Pascal Vincent
    Conference on Language Modeling (COLM), 2025.
    #NLP, #DL
    [arXiv]

  4. Boosting LLM Reasoning via Spontaneous Self-Correction
    , Tengyu Xu, Xuewei Wang, Zhengxing Chen, Di Jin, Liang Tan, Yen-Ting, Zishun Yu, Zhuokai Zhao, Yun He, Sinong Wang, Han Fang, et Chen Zhu
    Conference on Language Modeling (COLM), 2025.
    #NLP, #RL
    [openreview], [arXiv]

  5. Do Biased Models Have Biased Thoughts?
    Swati Rajwal, Shivank Garg, Reem Abdel-Salam et
    Conference on Language Modeling (COLM), 2025.
    #NLP
    [openreview], [arXiv]

  6. Revisiting Replay and Gradient Alignment for Continual Pre-Training of Large Language Models
    , Gopeshh Subbaraj, Matthew Riemer, Nizar Islah, Tsuguchika Tabaru, Hiroaki Kingetsu, et Irina Rish
    Conference on Lifelong Learning Agents (CoLLAs), 2025.
    #NLP, #DL
    [arXiv]

  7. Combining Domain and Alignment Vectors Provides Better Knowledge-Safety Trade-offs in LLMs
    , Quentin Fournier, Matthew Riemer, Pin-Yu Chen, Amal Zouaq, Payel Das et
    Annual Meeting of the Association for Computational Linguistics (ACL), 2025.
    #NLP
    [acl], [arXiv]

  8. Small Encoders Can Rival Large Decoders in Detecting Groundedness
    , , Quentin Fournier, Fernando Rodriguez, Alaa Boukhary, Adam Elwood et
    Findings of the Association for Computational Linguistics (ACL), 2025.
    #NLP
    [acl], [arXiv]

  9. Do Robot Snakes Dream like Electric Sheep? Investigating the Effects of Architectural Inductive Biases on Hallucination
    , Prasanna Parthasarathi, Mehdi Rezagholizadeh, Boxing Chen et
    Findings of the Association for Computational Linguistics (ACL), 2025.
    #NLP
    [acl], [arXiv]

  10. IDAT: A Multi-Modal Dataset and Toolkit for Building and Evaluating Interactive Task-Solving Agents
    Shrestha Mohanty, Negar Arabzadeh, Andrea Tupini, Yuxuan Sun, Alexey Skrynnik, , Marc-Alexandre Côté et Julia Kiseleva
    ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR), 2025.
    #NLP
    [arXiv]

  11. Compression via Pre-trained Transformers: A Study on Byte-Level Multimodal Data
    , Anian Ruoss, Joel Veness et Tim Genewein
    International Conference on Machine Learning, 2025.
    #NLP, #RL
    [openreview], [arXiv]

  12. NeoBERT: A Next Generation BERT
    , Quentin Fournier, Mariam El Mezouar et
    Transactions on Machine Learning Research (TMLR), 2025.
    #NLP
    [openreview]

  13. ChartGemma: Visual Instruction-tuning for Chart Reasoning in the Wild
    Ahmed Masry*, , Aayush Bajaj, Aaryaman Kartha, Enamul Hoque et Shafiq Joty
    International Conference on Computational Linguistics (COLING) Industry Track, 2025.
    #NLP
    [acl], [arXiv], [code]

2024

  1. WorkArena++: Towards Compositional Planning and Reasoning-based Common Knowledge Work Tasks
    Leo Boisvert*, , Maxime Gasse, Massimo Caccia, Thibault Le Sellier De Chezelles, Quentin Cappart, Nicolas Chapados, Alexandre Lacoste et Alexandre Drouin
    Conference on Neural Information Processing Systems (NeurIPS) Datasets and Benchmarks Track, 2024.
    #NLP
    [neurips], [openreview], [arXiv], [code]

  2. Exploring Quantization for Efficient Pre-Training of Transformer Language Models
    , Quentin Fournier, et
    Findings of the Association for Computational Linguistics (EMNLP), 2024.
    #NLP, #DL
    [acl], [arXiv]

  3. Context-Aware Assistant Selection for Improved Inference Acceleration with Large Language Models
    , Prasanna Parthasarathi, Mehdi Rezagholizadeh et
    Conference on Empirical Methods in Natural Language Processing (EMNLP), 2024.
    #NLP
    [acl], [arXiv]

  4. Do Large Language Models Know How Much They Know?
    , , Prasanna Parthasarathi, Shagun Sodhani et
    Conference on Empirical Methods in Natural Language Processing (EMNLP), 2024.
    #NLP
    [acl], [arXiv]

  5. Should We Attend More or Less? Modulating Attention for Fairness
    , , Samira Shabanian et
    Conference on Language Modeling (COLM), 2024.
    #NLP
    [openreview], [arXiv]

  6. Are self-explanations from Large Language Models faithful?
    , et Siva Reddy
    Findings of the Association for Computational Linguistics (ACL), 2024.
    #NLP
    [acl], [arXiv], [code], [YouTube]

  7. A deep-dive into the tradeoffs of preference alignment with PEFT
    , Quentin Fournier, Matthew Riemer, Pin-Yu Chen, Amal Zouaq, Payel Das et
    Annual Meeting of the Association for Computational Linguistics (ACL), 2024.
    #NLP
    [acl], [arXiv]

  8. Why Don’t Prompt-Based Fairness Metrics Correlate?
    , , Ioana Baldini et
    Annual Meeting of the Association for Computational Linguistics (ACL), 2024.
    #NLP
    [acl], [arXiv], [YouTube]

  9. Sub-goal Distillation: A Method to Improve Small Language Agents
    , Elias Stengel-Eskin, et Marc-Alexandre Cote
    Conference on Lifelong Learning Agents (CoLLAs), 2024. [Oral presentation.]
    #RL, #NLP
    [arXiv]

  10. Faithfulness Measurable Masked Language Models
    , Siva Reddy et
    International Conference on Machine Learning (ICML), 2024. [Spotlight award - top 3.5%]
    #NLP
    [pmlr], [arXiv], [code], [YouTube], [blogpost]

  11. MVP: Minimal Viable Phrase for Long Text Understanding
    , Amal Zouaq et
    Joint International Conference on Computational Linguistics, Language, Resources and Evaluation (LREC-COLING), 2024.
    #NLP
    [acl]

  12. Fairness-Aware Structured Pruning in Transformers
    , , Samira Shabanian, Ioana Baldini et
    AAAI Conference on Artificial Intelligence (AAAI), 2024.
    #NLP
    [aaai], [arXiv], [YouTube]

2023

  1. Self-Influence Guided Data Reweighting for Language Model Pre-training
    , Tolga Bolukbasi, Sriram Ganapathy, Shikhar Vashishth, et Partha Talukdar
    Conference on Empirical Methods in Natural Language Processing (EMNLP), 2023.
    #NLP
    [acl], [openreview], [arXiv]

  2. EpiK-Eval: Evaluation for Language Models as Epistemic Models
    , , Prasanna Parthasarathi, Shagun Sodhani et
    Conference on Empirical Methods in Natural Language Processing (EMNLP), 2023.
    #NLP
    [acl], [openreview], [arXiv], [code]

  3. Measuring the Knowledge Acquisition-Utilization Gap in Pretrained Language Models
    Amirhossein Kazemnejad, Mehdi Rezagholizadeh, Prasanna Parthasarathi et
    Findings of the Association for Computational Linguistics (EMNLP), 2023.
    #NLP
    [acl], [openreview], [arXiv]

  4. Deep Learning on a Healthy Data Diet: Finding Important Examples for Fairness
    , Prasanna Parthasarathi, , Hamid Palangi, Samira Shabanian et
    AAAI Conference on Artificial Intelligence (AAAI), 2023.
    #NLP
    [aaai], [arXiv], [YouTube]

2022

  1. Evaluating the Faithfulness of Importance Measures in NLP by Recursively Masking Allegedly Important Tokens and Retraining
    , Nicholas Meade, Vaibhav Adlakha et Siva Reddy
    Findings of the Association for Computational Linguistics (EMNLP), 2022.
    [BlackboxNLP, 2022]
    #NLP
    [acl], [arXiv], [code]

  2. Detecting Languages Unintelligible to Multilingual Models through Local Structure Probes
    , Prasanna Parthasarathi, Amal Zouaq et
    Findings of the Association for Computational Linguistics (EMNLP), 2022.
    #NLP
    [acl]

  3. Local Structure Matters Most in Most Languages
    , Prasanna Parthasarathi, Amal Zouaq et
    Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics and the International Joint Conference on Natural Language Processing (AACL-IJCNLP), 2022.
    #NLP
    [acl]

  4. Post-hoc Interpretability for Neural NLP: A Survey
    , Siva Reddy et
    ACM Computing Surveys, 2022.
    #NLP
    [acm], [arXiv]

  5. Local Structure Matters Most: Perturbation Study in NLU
    , Prasanna Parthasarathi, Amal Zouaq et
    Findings of the Association for Computational Linguistics (ACL), 2022.
    #NLP
    [acl], [arXiv]

2021

  1. Benchmarking Bias Mitigation Algorithms in Representation Learning through Fairness Metrics
    , Deepak Sharma, Soroush Mehri, Adriana Romero, Samira Shabanian et Sina Honari
    Conference on Neural Information Processing Systems (NeurIPS) Datasets and Benchmarks Track, 2021.
    #NLP
    [neurips], [openreview], [code]

  2. A Survey of Data Augmentation Approaches for NLP
    Steven Y. Feng, Varun Gangal, Jason Wei, , Soroush Vosoughi, Teruko Mitamura et Eduard Hovy
    Findings of the Association for Computational Linguistics (ACL-IJCNLP), 2021.
    #NLP
    [acl], [arXiv]

  3. MLMLM: Link Prediction with Mean Likelihood Masked Language Model
    , Philippe Trempe, Amal Zouaq et
    Findings of the Association for Computational Linguistics (ACL-IJCNLP), 2021.
    #NLP
    [acl], [arXiv]

  4. A Brief Study on the Effects of Training Generative Dialogue Models with a Semantic loss
    Prasanna Parthasarathi, , Joelle Pineau et
    Proceedings of the 22nd Annual SIGdial Meeting on Discourse and Dialogue, 2021.
    #NLP
    [acl]

  5. Do Encoder Representations of Generative Dialogue Models Encode Sufficient Information about the Task ?
    Prasanna Parthasarathi, et Joelle Pineau
    Proceedings of the 22nd Annual SIGdial Meeting on Discourse and Dialogue, 2021.
    #NLP
    [acl]