Research map / Large language models / Agents, code and math
Formalizing Learning from Language Feedback with Provable Guarantees
Wanqiao Xu, Allen Nie, Ruijie Zheng, Aditya Modi, Adith Swaminathan, Ching-An Cheng
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 HELiX hypothesis-sampling exploration instead of a single chain-of-thought hypothesis and action | Claude-Sonnet-3.5 v2 agent · Minesweeper | cumulative reward | Improved |
| use HELiX instead of a single CoT hypothesis and action | Claude-Sonnet-3.5 v2 agent · Battleship | cumulative reward | Improved |
| add the consensus-based exploitation step to optimistic exploration (ablation) | HELiX · Battleship and Minesweeper | cumulative reward | Improved |
Closest papers by meaning
- Reinforce LLM Reasoning through Multi-Agent Reflection (ICML 2025 · similarity 0.88)
- T1: Advancing Language Model Reasoning through Reinforcement Learning and Inference Scaling (ICML 2025 · similarity 0.88)
- Interpreting Learned Feedback Patterns in Large Language Models (NeurIPS 2024 · similarity 0.88)
- $\textbf{Re}^{2}$: Unlocking LLM Reasoning via Reinforcement Learning with Re-solving (ICLR 2026 · similarity 0.87)
- MA-RLHF: Reinforcement Learning from Human Feedback with Macro Actions (ICLR 2025 · similarity 0.87)
- STRIDE: Learnable Stepwise Language Feedback for LLM Reasoning (NeurIPS 2026 · similarity 0.87)
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.
