Publications

* = equal contribution.

Journal & Conference Papers

2026

• Expert-Aided Causal Discovery of Ancestral Graphs. Information Sciences
  Tiago da Silva, Bruna Bazaluk, Eliezer de Souza da Silva, António Góis, Salem Lahlou, Dominik Heider, Samuel Kaski, Diego Mesquita, Adèle Helena Ribeiro
• Particle GFlowNets: Rethinking Generative Marginalization Models. UAI 2026
  Tiago Silva, Diego Mesquita, Salem Lahlou
• Zero-Shot Off-Policy Learning. ICML 2026
  Arip Asadulaev, Maksim Bobrin, Salem Lahlou, Dmitry V. Dylov, Fakhri Karray, Martin Takáč
• Bayesian Rain Field Reconstruction using Commercial Microwave Links and Diffusion Model Priors. ICML 2026
  Badr Moufad, Albina Ilina, Hai Victor Habi, Salem Lahlou, Yazid Janati, Hagit Messer, Eric Moulines
• Path-dependent Discrete Amortized Inference (Oral). ICML 2026
  Tiago Silva, Nikolay Malkin, Salem Lahlou
• Avoid What You Know: Divergent Trajectory Balance for GFlowNets. ICML 2026 (+ ICLR 2026 DeLTa Workshop)
  Pedro Dall'Antonia, Tiago Silva, Daniel Csillag, Salem Lahlou, Diego Mesquita
• Noise-Corrected GRPO: From Noisy Rewards to Unbiased Gradients. ICML 2026
  Omar El Mansouri, Fathinah Asma Izzati, Mohamed El Amine Seddik, Salem Lahlou
• CoRe: Collaborative Reasoning via Cross Teaching. ICML 2026
  Kshitij Mishra, Mirat Aubakirov, Martin Takáč, Nils Lukas, Salem Lahlou
• Position: Agentic AI Systems Should Be Making Bayes-Consistent Decisions. ICML 2026
  Theodore Papamarkou, Pierre Alquier, Matthias Bauer, Wray Buntine, Andrew J. Davison, Gintare Karolina Dziugaite, Maurizio Filippone, Andrew Y. K. Foong, Vincent Fortuin, Dimitris Fouskakis, Jes Frellsen, Eyke Hüllermeier, Theofanis Karaletsos, Mohammad Emtiyaz Khan, Nikita Kotelevskii, Salem Lahlou, Yingzhen Li, Fang Liu, Clare Lyle, Thomas Möllenhoff, Konstantina Palla, Maxim Panov, Yusuf Sale, Kajetan Schweighofer, Artem Shelmanov, Siddharth Swaroop, Martin Trapp, Willem Waegeman, Andrew Gordon Wilson, Alexey Zaytsev
• FinChain: A Symbolic Benchmark for Verifiable Chain-of-Thought Financial Reasoning. ACL 2026
  Zhuohan Xie, Dhruv Sahnan, Debopriyo Banerjee, Georgi Georgiev, Rushil Thareja, Hachem Madmoun, Jinyan Su, Aaryamonvikram Singh, Yuxia Wang, Rui Xing, Fajri Koto, Haonan Li, Ivan Koychev, Tanmoy Chakraborty, Salem Lahlou, Veselin Stoyanov, Preslav Nakov
• SAHM: A Benchmark for Arabic Financial and Shariah-compliant Reasoning. ACL 2026
  Rania Elbadry, Sarfraz Ahmad*, Dani Bouch*, Momina Ahsan, Xueqing Peng, Jimin Huang, Muhra AlMahri, Marwa Elsaid Khalil, Yuxia Wang, Salem Lahlou, Veselin Stoyanov, Sophia Ananiadou, Preslav Nakov, Zhuohan Xie
• Temporal Contrastive Decoding: A Training-Free Method for Large Audio-Language Models. Findings of ACL 2026
  Yanda Li, Yuhan Liu, Zirui Song, Yunchao Wei, Martin Takáč, Salem Lahlou
• Byzantine Cheap Talk: Adversarial Resilience and Topology Effects in LLM Coordination Games. NETYS 2026
  Aya El Mir, Martin Takáč, Salem Lahlou
• Communication Enables Cooperation in LLM Agents: A Comparison with Curriculum-Based Approaches. EACL 2026 (Oral)
  Hachem Madmoun, Salem Lahlou
• SD-E2: Semantic Exploration for Reasoning Under Token Budgets. Findings of EACL 2026
  Kshitij Mishra, Nils Lukas, Salem Lahlou
• TorchGFN: A PyTorch GFlowNet Library. JMLR 2026 (+ NeurIPS 2025 Workshop on Frontiers in Probabilistic Inference)
  Joseph D. Viviano*, Omar G. Younis*, Sanghyeok Choi*, Victor Schmidt, Yoshua Bengio, Salem Lahlou

2025

• Loss-guided Auxiliary Agents for Overcoming Mode Collapse in GFlowNets. AAAI 2026
  Idriss Malek, Aya Laajil, Abhijith Sharma, Eric Moulines, Salem Lahlou
• Improved Exploration in GFlowNets via Enhanced Epistemic Neural Networks. ICoIAS 2025
  Sajan Muhammad, Salem Lahlou
• Assessing the Capabilities of Large Language Models for Oil and Gas Industry Applications. ADIPEC 2025
  F. Castanedo, W. Bhattacharya, S. Ghosh, M. Takac, S. Lahlou, Z. Iklassov, M. Schaffrath, N. Reddicharla, R. Mohan, M. Yaslam, M. Lee
• PORT: Preference Optimization on Reasoning Traces. NAACL 2025
  Salem Lahlou*, Abdalgader Abubaker*, Hakim Hacid
• On the Privacy Risks of Spiking Neural Networks: A Membership Inference Analysis. UAI 2025 (+ oral at ICLR 2025 Workshop ICBINB)
  Junyi Guan*, Abhijith Sharma*, Chong Tian, Salem Lahlou
• SVRPBench: A Realistic Benchmark for Stochastic Vehicle Routing Problem. NeurIPS 2025
  Ahmed Heakl*, Yahia Salaheldin Shaaban*, Martin Takac, Salem Lahlou, Zangir Iklassov

2023

• GFlowOut: Dropout with Generative Flow Networks. ICML 2023
  Dianbo Liu, Moksh Jain, Bonaventure F. P. Dossou, Qianli Shen, Salem Lahlou, Anirudh Goyal, Nikolay Malkin, Chris C. Emezue, Dinghuai Zhang, Nadhir Hassen, Xu Ji, Kenji Kawaguchi, Yoshua Bengio
• A Theory of Continuous Generative Flow Networks. ICML 2023
  Salem Lahlou, Tristan Deleu, Pablo Lemos, Dinghuai Zhang, Alexandra Volokhova, Alex Hernández-García, Léna Néhale Ezzine, Yoshua Bengio, Nikolay Malkin
• GFlowNet Foundations. JMLR 2023
  Yoshua Bengio*, Salem Lahlou*, Tristan Deleu*, Edward J. Hu, Mo Tiwari, Emmanuel Bengio
• GFlowNets and Variational Inference. ICLR 2023
  Nikolay Malkin*, Salem Lahlou*, Tristan Deleu*, Xu Ji, Edward Hu, Katie Everett, Dinghuai Zhang, Yoshua Bengio
• DEUP: Direct Epistemic Uncertainty Prediction. TMLR 2023
  Salem Lahlou*, Moksh Jain*, Hadi Nekoei, Victor Butoi, Paul Bertin, Jarrid Rector-Brooks, Maksym Korablyov, Yoshua Bengio
• MiniGrid & MiniWorld: Modular & Customizable Reinforcement Learning Environments for Goal-Oriented Tasks. NeurIPS 2023
  Maxime Chevalier-Boisvert, Bolun Dai, Mark Towers, Rodrigo Perez-Vicente, Lucas Willems, Salem Lahlou, Suman Pal, Pablo Samuel Castro, Jordan Terry

2019

• BabyAI: A Platform to Study the Sample Efficiency of Grounded Language Learning. ICLR 2019
  Maxime Chevalier-Boisvert*, Dzmitry Bahdanau*, Salem Lahlou, Lucas Willems, Chitwan Saharia, Thien Huu Nguyen, Yoshua Bengio

2017

• A Nash Equilibrium Formulation of a Tradable Credits Scheme for Incentivizing Transport Choices. Transportation Research Part B 2017
  Salem Lahlou, Laura Wynter

Workshop Papers & Extended Abstracts

2026

• Random Features for Discrete Amortized Samplers. LXAI Workshop at ICML 2026
  Tiago Silva, Amauri H Souza, Salem Lahlou
• Breaking the Martingale Curse: Multi-Agent Debate via Asymmetric Cognitive Potential Energy. ICML 2026 Workshop AgenticUQ, Under review at NeurIPS 2026
  Yuhan Liu, Juntian Zhang, Yichen Wu, Martin Takáč, Salem Lahlou, Xiuying Chen, Nils Lukas
• HyperKKL: Enabling Non-Autonomous State Estimation through Dynamic Weight Conditioning. AI&PDE Workshop at ICLR 2026
  Yahia Salaheldin Shaaban, Salem Lahlou, Abdelrahman Sayed Sayed

2025

• Curriculum-Augmented GFlowNets for mRNA Sequence Generation. NeurIPS 2025 Machine Learning in Structural Biology Workshop
  Aya Laajil, Abduragim Shtanchaev, Aya El Mir, Sajan Muhammad, Eric Moulines, Tiago Silva, Salem Lahlou
• Mitigating Societal Cognitive Overload in the Age of AI: Challenges and Directions. ICLR 2025 Workshop on Bidirectional Human-AI Alignment (to appear in PMLR)
  Salem Lahlou
• Customer Reactions to Companion AI: Exploring Opportunities and Threats for Vulnerable Consumers. Frontiers in Service 2025
  Salma André, Salem Lahlou, Margherita Pagani

2023

• BatchGFlowNet: Generative Flow Networks for Batch Active Learning. ICML 2023 SPIGM Workshop
  Shreshth A. Malik, Salem Lahlou, Andrew Jesson, Moksh Jain, Nikolay Malkin, Tristan Deleu, Yoshua Bengio, Yarin Gal

Preprints & Under Review

2026

• PETRA: Transforming Web Text for Petroleum-Engineering Domain Adaptation. Under review at EMNLP 2026 Industry Track
  Kirill Dubovikov, Omar El Mansouri, Hachem Madmoun, Yanda Li, Sandeep Kumar, Aya El Mir, Supriyo Ghosh, Writabrata Bhattacharya, Adrian Garcia-Garcia, Onkar Pandit, Sunil Kumar Sahu, Federico Castanedo, Larry Murray, Martin Takáč, Salem Lahlou
• Prompting Is Not Enough: Defining Quality in Synthetic QA Generation for Technical Domains. Under review at EMNLP 2026 Industry Track
  Aya El Mir, Adrian Garcia-Garcia, Kirill Dubovikov, Sunil Kumar Sahu, Federico Castanedo, Larry Murray, Martin Takáč, Salem Lahlou
• SCIRA: Pair-Free Defense Against Indirect Prompt Injection. Under review at AAAI 2027
  Haoyi Yang, Nils Lukas, Salem Lahlou, Kshitij Mishra
• PERSUADE: A Debate Arena for Benchmarking, Optimizing, and Auditing Persuasion in Language Models. Under review at AAAI 2027
  Sarim Hashmi, Samuele Poppi, Salem Lahlou, Martin Takáč, Kshitij Mishra, Nils Lukas
• CopyShield: A Cross-Level Benchmark of Copyright Defenses in LLMs. Under review at EMNLP 2026
  Maryam Alshehyari, Dushyant Singh Chauhan, Samuele Poppi, Martin Takáč, Salem Lahlou, Nils Lukas
• TRACER: Trace-Regularized Training Against Indirect Prompt Injection. Under review at EMNLP 2026
  Kshitij Mishra, Nils Lukas, Salem Lahlou
• AlphaEvo: LLM-Seeded Evolutionary Discovery of Robust Quantitative Trading Signals. Under review at NeurIPS 2026
  Temiko Machavariani, Kyriakos Tsourekas, Matvei Rotte, Luka Tsulaia, Iskhak Tazhibaev, Munachiso Samuel Nwadike, Salem Lahlou, Prashant Kumar Jamwal, Kentaro Inui, Martin Takáč, Zangir Iklassov
• Entropy-Gated Latent Recursion. Under review at NeurIPS 2026
  Soham Bhattacharjee, Dushyant Singh Chauhan, Salem Lahlou, Martin Takáč, Nils Lukas
• CausalDreamer: Learning Predictive World Models with Latent Disentanglement. Under review at NeurIPS 2026
  Prince Jha, Abhijith Sharma, Fan Feng, Kun Zhang, Salem Lahlou, Nils Lukas
• PROBE: Learning to Audit Policy Compliance in Tool-Using LLM Agents. Under review at NeurIPS 2026
  Kshitij Mishra, Abhijith Sharma, Nils Lukas, Salem Lahlou
• Bayesian Experimental Design: Bridging Optimization and Causal Discovery. Work In Progress
  Dhruv Goyal, Salem Lahlou

2025

• Accurate and Diverse LLM Mathematical Reasoning via Automated PRM-Guided GFlowNets. arXiv:2504.19981
  Adam Younsi*, Ahmed Attia*, Abdalgader Abubaker, Mohamed El Amine Seddik, Hakim Hacid, Salem Lahlou
• LLM-BABYBENCH: Understanding and Evaluating Grounded Planning and Reasoning in LLMs. arXiv:2505.12135
  Omar Choukrani*, Idriss Malek*, Daniil Orel, Zhuohan Xie, Zangir Iklassov, Martin Takáč, Salem Lahlou
• Semantic Communication Meets System 2 ML: How Abstraction, Compositionality and Emergent Languages Shape Intelligence. arXiv:2505.20964
  Mehdi Bennis, Salem Lahlou

2024

• On Generalization for Generative Flow Networks. arXiv:2407.03105
  Anas Krichel, Nikolay Malkin, Salem Lahlou, Yoshua Bengio
• Investigating Regularization of Self-Play Language Models. arXiv:2404.04291
  Reda Alami, Abdalgader Abubaker, Mastane Achab, Mohamed El Amine Seddik, Salem Lahlou

2020

• Mastering Rate-Based Curriculum Learning. arXiv preprint arXiv:2008.06456
  Lucas Willems*, Salem Lahlou*, Yoshua Bengio

Thesis and Edited Volumes

• Networked Systems: 13th International Conference, NETYS 2025, Rabat, Morocco, Proceedings. Lecture Notes in Computer Science, vol. 15736, Springer, 2025
  Salem Lahlou, Madhavan Mukund (Editors.)
• Advances in Uncertainty Modelling: From Epistemic Uncertainty Estimation to Generalized Generative Flow Networks. Ph.D. Thesis, Université de Montréal, 2023
  Salem Lahlou