Research map / Reinforcement learning / Visual and representation RL
Explicable Policy Search
Ze Gong, Yu Zhang
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 EPS instead of SAC | SAC · Navigation D1 | explicability score: 92.57 -> 94.87 | Improved |
| use EPS instead of SAC | SAC · Navigation D3 | average return: 94.47 -> 96.80 | Improved |
| use EPS instead of SAC | SAC · Navigation D1 | average return: 97.46 -> 95.53 | Hurt |
| use EPS instead of SAC | SAC · Navigation D4 | average return: 93.27 -> 92.37 | Hurt |
| use EPS surrogate reward plus engineered reward instead of preference-only reward (DRLHP) | DRLHP · Navigation D2 | average return: 23.28 -> 94.07 | Improved |
| use EPS instead of DRLHP | DRLHP · Navigation D3 | explicability score: -50.79 -> 94.72 | Improved |
| use EPS instead of policy shaping (PS) | PS · Navigation D3 | average return: -178.76 -> 96.80 | Improved |
| use EPS instead of policy shaping (PS) | PS · Navigation D2 | average return: 95.55 -> 94.07 | No clear effect |
| use EPS instead of policy shaping (PS) | PS · Navigation D1 | explicability score: 89.17 -> 94.87 | Improved |
| use EPS instead of policy shaping (PS) | PS · Navigation D4 | explicability score: 90.11 -> 92.17 | Improved |
| use EPS instead of SAC | SAC · Autonomous driving user study | mean human rating: 5.0 -> 7.6 | Improved |
| use EPS instead of DRLHP | DRLHP · Autonomous driving user study | mean human rating: 5.9 -> 7.6 | Improved |
Closest papers by meaning
- Beyond Single Stationary Policies: Meta-Task Players as Naturally Superior Collaborators (NeurIPS 2024 · similarity 0.86)
- Learning to Collaborate with Unknown Agents in the Absence of Reward (AAAI 2025 · similarity 0.86)
- NestRL: A Nested Training Regime for Mutual Adaptation in Human–AI Teaming (NeurIPS 2026 · similarity 0.86)
- Discovering Creative Behaviors through DUPLEX: Diverse Universal Features for Policy Exploration (NeurIPS 2024 · similarity 0.86)
- DUO: Diverse, Uncertain, On-Policy Query Generation and Selection for Reinforcement Learning from Human Feedback (AAAI 2025 · similarity 0.85)
- Expectation Alignment: Handling Reward Misspecification in the Presence of Expectation Mismatch (NeurIPS 2024 · similarity 0.85)
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.
