Research map / Reinforcement learning / RL from feedback
Belief-Driven Value Alignment for Human-Robot Collaboration
Saisai Li, Bing Shi, Yiming Xia, Xiao Su
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 |
|---|---|---|---|
| raise human rationality coefficient beta from 1 to 5 (ablation) | PFHDP · ChefWorld | convergence step of belief alignment: 260 -> 140 | Improved |
| raise human rationality coefficient beta from 1 to 10 (ablation) | PFHDP · ChefWorld | convergence step of belief alignment: 260 -> 80 | Improved |
| raise human rationality coefficient beta from 5 to 10 (ablation) | PFHDP · ChefWorld | degree of true value alignment bR(theta) | Hurt |
| lower beta to 0.5 (near-random human) (ablation) | PFHDP · ChefWorld | belief alignment convergence | Hurt |
| add human estimate of robot belief (PFHDP vs ECIRL-R) | ECIRL-R · ChefWorld and GridWorld | belief alignment and mean cumulative reward | Improved |
| model human as boundedly rational instead of perfectly rational (ECIRL-R vs ECIRL) | ECIRL · ChefWorld | belief in true preference | Improved |
| increase environment scale | all algorithms · GridWorld | mean cumulative reward | Hurt |
Closest papers by meaning
- An Evolutionary Perspective on AI Alignment (Student Abstract) (AAAI 2025 · similarity 0.89)
- DUO: Diverse, Uncertain, On-Policy Query Generation and Selection for Reinforcement Learning from Human Feedback (AAAI 2025 · similarity 0.86)
- Robust Reward Alignment via Hypothesis Space Batch Cutting (ICML 2025 · similarity 0.86)
- PARL: A Unified Framework for Policy Alignment in Reinforcement Learning from Human Feedback (ICLR 2024 · similarity 0.86)
- Personalizing Reinforcement Learning from Human Feedback with Variational Preference Learning (NeurIPS 2024 · similarity 0.86)
- Safe RLHF: Safe Reinforcement Learning from Human Feedback (ICLR 2024 · 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.
