Marc G. Bellemarehttps://marcgbellemare.info/en/Recent content on Marc G. BellemareHugoen© {year} Marc G. BellemareThu, 01 Jan 2026 00:00:00 +0000Compositional Planning with Jumpy World Modelshttps://marcgbellemare.info/en/publications/farebrother26compositional/Thu, 01 Jan 2026 00:00:00 +0000https://marcgbellemare.info/en/publications/farebrother26compositional/Convergence Theorems for Entropy-Regularized and Distributional Reinforcement Learninghttps://marcgbellemare.info/en/publications/jhaveri25convergence/Tue, 01 Jul 2025 00:00:00 +0000https://marcgbellemare.info/en/publications/jhaveri25convergence/Tapered Off-Policy REINFORCE: Stable and Efficient Reinforcement Learning for LLMshttps://marcgbellemare.info/en/publications/leroux25tapered/Tue, 01 Jul 2025 00:00:00 +0000https://marcgbellemare.info/en/publications/leroux25tapered/A Distributional Analogue to the Successor Representationhttps://marcgbellemare.info/en/publications/wiltzer24successor/Mon, 01 Jul 2024 00:00:00 +0000https://marcgbellemare.info/en/publications/wiltzer24successor/Action Gaps and Advantages in Continuous-Time Distributional Reinforcement Learninghttps://marcgbellemare.info/en/publications/wiltzer24continuous/Mon, 01 Jul 2024 00:00:00 +0000https://marcgbellemare.info/en/publications/wiltzer24continuous/An Analysis of Quantile Temporal-Difference Learninghttps://marcgbellemare.info/en/publications/rowland24quantile/Mon, 01 Jul 2024 00:00:00 +0000https://marcgbellemare.info/en/publications/rowland24quantile/Controlling Large Language Model Agents with Entropic Activation Steeringhttps://marcgbellemare.info/en/publications/rahn24steering/Mon, 01 Jul 2024 00:00:00 +0000https://marcgbellemare.info/en/publications/rahn24steering/A Novel Stochastic Gradient Descent Algorithm for Learning Principal Subspaceshttps://marcgbellemare.info/en/publications/lelan23subspaces/Sat, 01 Jul 2023 00:00:00 +0000https://marcgbellemare.info/en/publications/lelan23subspaces/Bigger, Better, Faster: Human-level Atari with Human-Level Efficiencyhttps://marcgbellemare.info/en/publications/schwarzer23bigger/Sat, 01 Jul 2023 00:00:00 +0000https://marcgbellemare.info/en/publications/schwarzer23bigger/Bootstrapped Representations in Reinforcement Learninghttps://marcgbellemare.info/en/publications/lelan23bootstrapped/Sat, 01 Jul 2023 00:00:00 +0000https://marcgbellemare.info/en/publications/lelan23bootstrapped/Discovering the Electron Beam Induced Transition Rates for Silicon Dopants in Graphene with Deep Neural Networks in the STEMhttps://marcgbellemare.info/en/publications/roccapriore23electron/Sat, 01 Jul 2023 00:00:00 +0000https://marcgbellemare.info/en/publications/roccapriore23electron/Investigating Multi-Task Pretraining and Generalization in Reinforcement Learninghttps://marcgbellemare.info/en/publications/taiga23investigating/Sat, 01 Jul 2023 00:00:00 +0000https://marcgbellemare.info/en/publications/taiga23investigating/Policy Optimization in a Noisy Neighborhood: On Return Landscapes in Continuous Controlhttps://marcgbellemare.info/en/publications/rahn23policy/Sat, 01 Jul 2023 00:00:00 +0000https://marcgbellemare.info/en/publications/rahn23policy/Proto-Value Networks: Scaling Representation Learning with Auxiliary Taskshttps://marcgbellemare.info/en/publications/farebrother23proto/Sat, 01 Jul 2023 00:00:00 +0000https://marcgbellemare.info/en/publications/farebrother23proto/Sample-Efficient Reinforcement Learning by Breaking the Replay Ratio Barrierhttps://marcgbellemare.info/en/publications/doro23sample/Sat, 01 Jul 2023 00:00:00 +0000https://marcgbellemare.info/en/publications/doro23sample/Small Batch Deep Reinforcement Learninghttps://marcgbellemare.info/en/publications/obandoceron23small/Sat, 01 Jul 2023 00:00:00 +0000https://marcgbellemare.info/en/publications/obandoceron23small/The Statistical Benefits of Quantile Temporal-Difference Learning for Value Estimationhttps://marcgbellemare.info/en/publications/rowland23benefits/Sat, 01 Jul 2023 00:00:00 +0000https://marcgbellemare.info/en/publications/rowland23benefits/Distributional Reinforcement Learninghttps://marcgbellemare.info/en/books/distributional-rl/Tue, 30 May 2023 00:00:00 +0000https://marcgbellemare.info/en/books/distributional-rl/<p><strong>Marc G. Bellemare, Will Dabney, Mark Rowland</strong> MIT Press, May 2023 · 384 pages · Adaptive Computation and Machine Learning series</p> <ul> <li><a href="https://mitpress.mit.edu/9780262048019/distributional-reinforcement-learning/">MIT Press page</a></li> <li><a href="https://www.distributional-rl.org/">Book website</a></li> <li><a href="https://direct.mit.edu/books/oa-monograph/5590/Distributional-Reinforcement-Learning">Open-access PDF</a></li> <li>ISBN (hardcover): 9780262048019 · ISBN (eBook): 9780262374019</li> </ul> <hr> <p>This is the first comprehensive guide to distributional reinforcement learning, providing a new mathematical formalism for thinking about decisions from a probabilistic perspective. Rather than computing expected values, the book focuses on how total reward behaves as a probability distribution — presenting core concepts, mathematical proofs, and algorithmic developments for characterising, computing, estimating, and making decisions based on random returns. Applications span finance, computational neuroscience, psychology, macroeconomics, and robotics.</p>Distributional Reinforcement Learninghttps://marcgbellemare.info/en/publications/bellemare23book/Sat, 01 Apr 2023 00:00:00 +0000https://marcgbellemare.info/en/publications/bellemare23book/Distributional Hamilton-Jacobi-Bellman Equations for Continuous-Time Reinforcement Learninghttps://marcgbellemare.info/en/publications/wiltzer22hjb/Fri, 01 Jul 2022 00:00:00 +0000https://marcgbellemare.info/en/publications/wiltzer22hjb/