Research map / Reinforcement learning / Policy gradient and actor-critic
Batch Value-function Approximation with Only Realizability
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 |
|---|---|---|---|
| Use pairwise comparisons over candidate value functions, with a partition built from each compared pair | BVFT for batch RL · Batch RL with realizability and the paper's exploratory-data assumption | sample complexity: \tilde{O}\left(\frac{C^2 \ln\frac{|\mathcal{F}|}{\delta}}{\epsilon^4 (1-\gamma)^8}\right) | Helped |
| Use a piecewise-constant class to validate candidate value functions through a projected Bellman update | BVFT · Batch RL with realizability | policy suboptimality bound: \frac{(4+8\sqrt{C})\epsilon_{\mathcal{F}}}{(1-\gamma)^2} + \epsilon\cdot V_{\max} | Helped |
Closest papers by meaning
- Computational Hardness of Reinforcement Learning with Partial $q^{\pi}$-Realizability (NeurIPS 2025 · similarity 0.89)
- Switching the Loss Reduces the Cost in Batch Reinforcement Learning (ICML 2024 · similarity 0.89)
- A Nearly Optimal and Low-Switching Algorithm for Reinforcement Learning with General Function Approximation (NeurIPS 2024 · similarity 0.89)
- Sample-Efficiency in Multi-Batch Reinforcement Learning: The Need for Dimension-Dependent Adaptivity (ICLR 2024 · similarity 0.88)
- Reward-Free Kernel-Based Reinforcement Learning (ICML 2024 · similarity 0.88)
- Oracle-Efficient Reinforcement Learning for Max Value Ensembles (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.