Research map / Reinforcement learning / MDPs and planning
Towards a Sharp Analysis of Learning Offline $f$-Divergence-Regularized Contextual Bandits
Qingyue Zhao, Kaixuan Ji, Heyang Zhao, Tong Zhang, Quanquan Gu
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 |
|---|---|---|---|
| use reverse-KL regularization with less-covered reference policy (larger concentrability) | KL-PCB · linear bandit (d=20, 100 arms) | sub-optimality gap | Hurt |
| use chi-square regularization (strongly convex f) with varying coverage | f-CB · linear bandit (d=20, 100 arms) | sub-optimality gap | No clear effect |
| use spiky instead of uniform reference policy under KL regularization | KL-PCB · MNIST contextual bandit | sub-optimality gap | Hurt |
| use spiky instead of uniform reference policy under chi-square regularization | f-CB · MNIST contextual bandit | sub-optimality gap | No clear effect |
| scale strong convexity modulus alpha as 2^15/n | f-CB · two-armed bandit | sub-optimality | No clear effect |
Closest papers by meaning
- Fast Rates for Offline Contextual Bandits with Forward-KL Regularization under Single-Policy Concentrability (NeurIPS 2026 · similarity 0.93)
- Sharp Analysis for KL-Regularized Contextual Bandits and RLHF (NeurIPS 2025 · similarity 0.90)
- On the Optimal Sample Complexity of Offline Multi-Armed Bandits with KL Regularization (NeurIPS 2026 · similarity 0.90)
- Worst-Case Offline Reinforcement Learning with Arbitrary Data Support (NeurIPS 2024 · similarity 0.88)
- Relaxed Stationary Distribution Correction Estimation for Improved Offline Policy Optimization (AAAI 2024 · similarity 0.87)
- Robust Offline Reinforcement Learning with Linearly Structured $f$-Divergence Regularization (ICML 2025 · similarity 0.87)
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.
