Research map / Reinforcement learning / Offline RL
Optimizing Interventions via Offline Policy Evaluation: Studies in Citizen Science
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 TCI optimized corrected multi-message policy instead of myopic helpful-message-at-predicted-disengagement policy | TCI · Galaxy Zoo live study 1 | user contributions: 69% more contributions | Helped |
| add policy correction step (Step 3) to the target policy (ablation) | TCI · Galaxy Zoo live study 2, 1-session users | contribution rate | Helped |
| random-timing interventions with all message types versus no intervention | Galaxy Zoo messages · Galaxy Zoo live study 1 | user contributions | No clear effect |
| add prediction-based timing and policy optimization over messages versus random intervention | Intervention messages · Galaxy Zoo live study 1 | user contributions | Helped |
| importance sampling policy value estimate in offline evaluation | Importance sampling estimator · Galaxy Zoo simulator | absolute error of value estimate | Hurt |
Closest papers by meaning
- Efficient Policy Evaluation with Offline Data Informed Behavior Policy Design (ICML 2024 · similarity 0.83)
- Simple Ingredients for Offline Reinforcement Learning (ICML 2024 · similarity 0.83)
- STITCH-OPE: Trajectory Stitching with Guided Diffusion for Off-Policy Evaluation (NeurIPS 2025 · similarity 0.83)
- A Snapshot of Influence: A Local Data Attribution Framework for Online Reinforcement Learning (NeurIPS 2025 · similarity 0.83)
- Iteratively Refined Behavior Regularization for Offline Reinforcement Learning (NeurIPS 2024 · similarity 0.83)
- The Generalization Gap in Offline Reinforcement Learning (ICLR 2024 · similarity 0.82)
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.