mlsanityData to 14 Oct 2026

Learning Perception to Action Rules through Reinforcement Learning for Theory-of-Mind Reasoning

arXiv 2610.19796 ↗First posted 14 Oct 2026cs.AI cs.CLcited by 0

On the map: LLM reasoning → Reinforcement learning for reasoning

Abstract (arXiv)

In Theory-of-Mind (ToM) reasoning, large language models may correctly infer the agent's mental state but fail to use that inference when predicting or judging the agent's subsequent actions. In this work, we investigate whether explicitly connecting perception, belief, and action with rules improves ToM reasoning. We generate concise textitif--then rules that express relations between perception to action and use Group Relative Policy Optimization (GRPO) to train Qwen2.5 and Llama3 models across four ToM benchmarks. GRPO with rule generation outperforms prompting, supervised fine-tuning, and think reasoning. Moreover, rules generated by larger models (≥3B) provide transferable knowledge that improves downstream performance of GPT-4o-mini. These findings suggest that explicit rules help models bridge the reasoning gap in ToM and support knowledge transfer across models.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Reinforcement learning for reasoning, one of 17 directions in LLM reasoning. Builds on 11 of its 19 references that are in the map. First posted on arXiv 14 Oct 2026.

Builds on

11 of 19 references.

Closest ancestors · 2

Same direction, newest first

  1. 2025TemplateRL: Structured Template-Guided Reinforcement Learning for LLM Reasoning21 May
  2. 2025DAPO: An Open-Source LLM Reinforcement Learning System at Scale18 Mar

Other references · 9

  1. 2022Measuring and Narrowing the Compositionality Gap in Language Modelscited by 1,075
  2. 2025Chain of Draft: Thinking Faster by Writing Lesscited by 297
  3. 2023Understanding Social Reasoning in Language Models with Language Modelscited by 172
  4. 2024The Benefits of a Concise Chain of Thought on Problem-Solving in Large Language Modelscited by 63
  5. 2024SimpleToM: Exposing the Gap between Explicit ToM Inference and Implicit ToM Application in LLMscited by 57
  6. 2025Towards Dynamic Theory of Mind: Evaluating LLM Adaptation to Temporal Evolution of Human Statescited by 19
  7. 2025Metis-SPECS: Decoupling Multimodal Learning via Self-distilled Preference-based Cold Startcited by 5
  8. 2026Correct, Concise and Complete: Multi-stage Training For Adaptive Reasoningcited by 4
  9. 2026From Shortcuts to Reasoning: Robust Post-Training of Theory of Mind with Reinforcement Learningcited by 2

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025Do Theory of Mind Benchmarks Need Explicit Human-like Reasoning in Language Models?
  2. 2025MindPower: Enabling Theory-of-Mind Reasoning in VLM-based Embodied Agents
  3. 2023Boosting Theory-of-Mind Performance in Large Language Models via Prompting
  4. 2026Understanding Artificial Theory of Mind: Perturbed Tasks and Reasoning in Large Language Models
  5. 2026All Roads Lead to Rome: Incentivizing Divergent Thinking in Vision-Language Models
  6. 2026Avalon-ToM-Bench: Evaluating Fine-Grained Theory of Mind via Asymmetric Game Mechanics
  7. 2025Think or Not? Selective Reasoning via Reinforcement Learning for Vision-Language Models
  8. 2025Theory of Mind in Action: The Instruction Inference Task in Dynamic Human-Agent Collaboration