Research map / Reinforcement learning / Offline RL
Offline Minimax Soft-Q-learning Under Realizability and Partial Coverage
Masatoshi Uehara, Nathan Kallus, Jason D. Lee, Wen Sun
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Replace a hard-margin requirement with the soft-margin condition | MQP · Offline RL | assumption stringency | Improved |
| Require coverage only for a single comparator policy rather than uniformly over policies | MSQP and MQP · Offline RL | coverage requirement | Improved |
| Use refined, function-class-adaptive concentrability instead of requiring a finite density ratio for the initial distribution | MSQP · Offline RL with initial-distribution shift | coverage applicability | Improved |
| Use minimax objectives designed for L2 estimation rather than Bellman-residual guarantees alone | MSQP and MQP · Offline RL with general function approximation | L2-error guarantee | Improved |
| Relax the Bellman equality constraint to an inequality in the minimum-norm formulation | MSQP and MQP · Offline RL | existence of a convex minimax formulation | Improved |
Closest papers by meaning
- On the Complexity of Offline Reinforcement Learning with Q*-Approximation and Partial Coverage (NeurIPS 2026 · similarity 0.91)
- Exclusively Penalized Q-learning for Offline Reinforcement Learning (NeurIPS 2024 · similarity 0.90)
- Minimax Optimal and Computationally Efficient Algorithms for Distributionally Robust Offline Reinforcement Learning (NeurIPS 2024 · similarity 0.88)
- Offline Constrained Reinforcement Learning under Partial Data Coverage (NeurIPS 2026 · similarity 0.88)
- Pessimistic Nonlinear Least-Squares Value Iteration for Offline Reinforcement Learning (ICLR 2024 · similarity 0.88)
- Worst-Case Offline Reinforcement Learning with Arbitrary Data Support (NeurIPS 2024 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
