Research map / Reinforcement learning / Policy gradient and actor-critic
State-Action Similarity-Based Representations for Off-Policy Evaluation
Brahma S Pavse, Josiah P. Hanna
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 |
|---|---|---|---|
| feed ROPE representations into FQE instead of raw state-actions | FQE · HalfCheetah-medium (D4RL) | RMAE: 0.770 -> 0.247 | Improved |
| feed ROPE representations into FQE instead of raw state-actions | FQE · HalfCheetah-medium-expert (D4RL) | RMAE: 0.374 -> 0.078 | Improved |
| feed ROPE representations into FQE instead of raw state-actions | FQE · HalfCheetah-random (D4RL) | RMAE: 0.807 -> 0.990 | Hurt |
| use ROPE representations where FQE diverges | FQE · Walker2D-medium (D4RL) | RMAE: Diverged -> 0.462 | Improved |
| use ROPE representations where FQE diverges | FQE · Hopper-medium (D4RL) | RMAE: Diverged -> 0.208 | Improved |
| ROPE encoder instead of BCRL representation | BCRL · Hopper-medium (D4RL) | RMAE: 61.223 -> 0.208 | Improved |
| ROPE encoder instead of BCRL representation | BCRL · Walker2D-medium-expert (D4RL) | RMAE: 0.962 -> 0.252 | Improved |
| ROPE vs BCRL on random dataset | BCRL · HalfCheetah-random (D4RL) | RMAE: 0.979 -> 0.990 | No clear effect |
| ROPE vs state-action metrics (on-policy MICO) for grouping | on-policy MICO · 3x3 gridworld | grouping aligned with q^pi_e | Improved |
| vary ROPE encoder output dimension (ablation) | ROPE · Walker2D-medium (D4RL) | RMAE | No clear effect |
| vary beta (angular distance weight) (ablation) | ROPE · Walker2D-medium (D4RL) | RMAE | No clear effect |
| double FQE network capacity (FQE-deep) (ablation) | FQE · performance profile over runs | fraction of runs with RMAE <= 10 | No clear effect |
Closest papers by meaning
- Off-Policy Evaluation via Off-Policy Classification (NeurIPS 2019 · similarity 0.90)
- Kernel Metric Learning for In-Sample Off-Policy Evaluation of Deterministic RL Policies (ICLR 2024 · similarity 0.89)
- Statistical Tractability of Off-policy Evaluation of History-dependent Policies in POMDPs (ICLR 2025 · similarity 0.88)
- Model Selection for Off-policy Evaluation: New Algorithms and Experimental Protocol (NeurIPS 2025 · similarity 0.88)
- Efficient Off-Policy Learning for High-Dimensional Action Spaces (ICLR 2025 · similarity 0.88)
- A Unifying View of Linear Function Approximation in Off-Policy RL Through Matrix Splitting and Preconditioning (NeurIPS 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.
