Research map / Reinforcement learning / RL from feedback
Off-Policy Evaluation for Human Feedback
Qitong Gao, Ge Gao, Juncheng Dong, Vahid Tarokh, Min Chi, Miroslav Pajic
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 OPEHF instead of Rescale baseline with IS | IS estimator · Intelligent tutoring | MAE: 0.77 -> 0.57 | Improved |
| use OPEHF instead of Rescale baseline with DICE | DICE estimator · Intelligent tutoring | MAE: 2.33 -> 1.01 | Improved |
| use OPEHF instead of Fusion baseline with FQE | FQE estimator · Intelligent tutoring | MAE: 0.74 -> 0.59 | Improved |
| use OPEHF instead of Rescale baseline with IS | IS estimator · Intelligent tutoring | Rank: 0.4 -> 0.8 | Improved |
| use OPEHF reconstruction instead of standalone VLM-H (ablation) | OPEHF · Intelligent tutoring | MAE: 1.00 -> 0.57 | Improved |
| use OPEHF instead of standalone VLM-H (ablation) | OPEHF with IS · Intelligent tutoring | Rank: 0.41 -> 0.8 | Improved |
| use OPEHF instead of standalone VLM-H (ablation) | OPEHF with IS · Intelligent tutoring | Regret@1: 0.28 -> 0.41 | No clear effect |
| use OPEHF instead of standalone VLM-H with DR (ablation) | OPEHF with DR · DBS Patient #0 | MAE: 0.97 -> 0.09 | Improved |
| use OPEHF instead of Rescale baseline with DICE (ablation) | DICE estimator · DBS Patient #1 | MAE: 0.45 -> 0.87 | Hurt |
| use OPEHF instead of Fusion baseline with DICE | DICE estimator · DBS Patient #3 | MAE: 0.41 -> 1.03 | Hurt |
| use OPEHF instead of Fusion baseline with FQE | FQE estimator · DBS Patient #2 | MAE: 1.6 -> 0.38 | Improved |
| use Rescale instead of Fusion for environmental and human returns | IS/DR/DICE/FQE baselines · DBS cohort | MAE | Improved |
Closest papers by meaning
- Benchmarks for Deep Off-Policy Evaluation (ICLR 2021 · similarity 0.90)
- OPERA: Automatic Offline Policy Evaluation with Re-weighted Aggregates of Multiple Estimators (NeurIPS 2024 · similarity 0.89)
- Model Selection for Off-policy Evaluation: New Algorithms and Experimental Protocol (NeurIPS 2025 · similarity 0.88)
- Abstract Reward Processes: Leveraging State Abstraction for Consistent Off-Policy Evaluation (NeurIPS 2024 · similarity 0.87)
- STITCH-OPE: Trajectory Stitching with Guided Diffusion for Off-Policy Evaluation (NeurIPS 2025 · similarity 0.87)
- Demystifying the Paradox of Importance Sampling with an Estimated History-Dependent Behavior Policy in Off-Policy Evaluation (ICML 2025 · similarity 0.86)
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.
