Research map / Reinforcement learning / MDPs and planning
Reinforcement Learning in Low-rank MDPs with Density Features
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Clip the regression target (recursively clipped occupancy) instead of using raw importance weights in off-policy occupancy estimation (FORC) | off-policy occupancy estimation in low-rank MDPs with density features · offline, arbitrary (non-exploratory) data | l1 estimation error ||d_hat - d||_1 | No clear effect |
| Build exploratory policy cover via barycentric spanner of estimated occupancies, level by level (FORCE) instead of a fixed exploration scheme | online reward-free exploration in low-rank MDPs · online, known density features | sample complexity to achieve l1 occupancy error epsilon | Helped |
| Joint linearization with a single shared feature across policies in representation learning (replacing per-policy linearization) | online occupancy estimation with unknown density features (ForcRlE) · online, representation learning | concentrability coefficient of policy cover | Helped |
Closest papers by meaning
- Shift Before You Learn: Enabling Low-Rank Representations in Reinforcement Learning (NeurIPS 2025 · similarity 0.89)
- Harnessing Density Ratios for Online Reinforcement Learning (ICLR 2024 · similarity 0.88)
- Occupancy-based Policy Gradient: Estimation, Convergence, and Optimality (NeurIPS 2024 · similarity 0.88)
- Learning from Sparse Offline Datasets via Conservative Density Estimation (ICLR 2024 · similarity 0.88)
- Optimal Single-Policy Sample Complexity and Transient Coverage for Average-Reward Offline RL (NeurIPS 2025 · similarity 0.88)
- Generative OOD-regularized Model-based Policy Optimization (NeurIPS 2026 · 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.