Research map / Reinforcement learning / Imitation learning
Policy Learning Using Weak Supervision
Jingkang Wang, Hongyi Guo, Zhaowei Zhu, Yang Liu
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 peer reward instead of noisy reward | DDQN · CartPole e=0.1 | N_epi: 114.6 -> 86.2 | Improved |
| use peer reward instead of noisy reward | DDQN · CartPole e=0.4 | N_epi: 163.1 -> 106.9 | Improved |
| use peer reward instead of noisy reward | DDQN · CartPole e=0.3 | R_avg: 177.1 -> 174.1 | Hurt |
| use peer reward instead of surrogate reward | DDQN · CartPole e=0.3 | R_avg: 192.2 -> 174.1 | Hurt |
| use peer reward instead of true reward | DDQN · CartPole e=0.1 | N_epi: 101.2 -> 86.2 | Improved |
| use peer reward instead of noisy reward | DDQN · CartPole e=0.2 | R_avg: 168.8 -> 195.5 | Improved |
| use PeerBC (xi = 0.2) instead of standard BC | BC · Atari Pong | score: 14.7 -> 18.8 | Improved |
| use PeerBC (xi = 0.5) instead of standard BC | BC · Atari Enduro | score: 138.9 -> 230.9 | Improved |
| use PeerBC (xi = 0.5) instead of standard BC | BC · Atari Boxing | score: 56.2 -> 75.6 | Improved |
| increase peer penalty xi from 0.5 to 1.0 (ablation) | PeerBC · Atari Freeway | score: 22.4 -> 8.9 | Hurt |
| increase peer penalty xi from 0.2 to 0.5 (ablation) | PeerBC · Atari lift over expert | lift: +11.3% -> +19.5% | Improved |
| use PeerBC instead of standard BC | BC · CartPole-v1 imitation learning | return: 18% | Improved |
Closest papers by meaning
- Asymmetric REINFORCE for off-Policy Reinforcement Learning: Balancing positive and negative rewards (NeurIPS 2025 · similarity 0.85)
- Peer Learning: Learning Complex Policies in Groups from Scratch via Action Recommendations (AAAI 2024 · similarity 0.85)
- Behavior Cloning is Not All You Need: The Optimality of On-Policy Distillation for Noisy Expert Feedback (NeurIPS 2026 · similarity 0.85)
- LLM-as-a-Tutor: Policy-Aware Prompt Adaptation for Non-Verifiable RL (NeurIPS 2026 · similarity 0.85)
- Skill-SD: Skill-Conditioned Self-Distillation for Multi-turn LLM Agents (NeurIPS 2026 · similarity 0.85)
- Hybrid Inverse Reinforcement Learning (ICML 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.
