Research map / Reinforcement learning / RL from feedback
Interactive Learning from Policy-Dependent Human Feedback
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 policy-dependent (advantage-based) feedback instead of policy-independent feedback | COACH vs TAMER and Q learning · Cliff world, improvement feedback | steps to goal | Helped |
| Set TAMER-style policy-independent feedback model (action feedback) for TAMER and COACH | TAMER and COACH · Cliff world, action feedback | steps to goal | No clear effect |
| Use COACH with lambda=0 trace (no eligibility trace) under sparse task feedback (ablation) | COACH · Cliff world, task feedback | steps to goal | Hurt |
| Add eligibility trace (lambda=0.9) to COACH under task feedback (ablation) | COACH · Cliff world, task feedback | steps to goal | Helped |
| Use TAMER with sparse task feedback | TAMER · Cliff world, task feedback | convergence | Hurt |
| Add eligibility traces and policy-dependent feedback in real-time robot training (COACH) vs TAMER | TurtleBot robot behaviors · TurtleBot five behaviors | behaviors learned | Helped |
| Use COACH's policy-dependent update where positive feedback reinforces staying (vs TAMER reward approximation) | TurtleBot stay/forward scenario · TurtleBot ball/cylinder scenario | qualitative response to positive feedback | Helped |
Closest papers by meaning
- Towards Better Robot Learners: Leveraging Implicit and Explicit Human Feedback Together in Human Robot Interactions (AAAI 2025 · similarity 0.89)
- DUO: Diverse, Uncertain, On-Policy Query Generation and Selection for Reinforcement Learning from Human Feedback (AAAI 2025 · similarity 0.88)
- GUIDE: Real-Time Human-Shaped Agents (NeurIPS 2024 · similarity 0.87)
- Personalizing Reinforcement Learning from Human Feedback with Variational Preference Learning (NeurIPS 2024 · similarity 0.86)
- A Minimaximalist Approach to Reinforcement Learning from Human Feedback (ICML 2024 · similarity 0.86)
- Reward Learning from Multiple Feedback Types (ICLR 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.