To Think or Not To Think, That is The Question for Large Reasoning Models in Theory of Mind Tasks

2026-02-11

AI Summary

This study investigates whether large language models genuinely possess theory of mind (ToM)—the capacity to infer others’ latent mental states such as beliefs and intentions—rather than relying on superficial cues. Through three established ToM benchmarks, the authors systematically evaluate nine state-of-the-art models, comparing reasoning-based and non-reasoning architectures while introducing fine-grained response analysis. Findings reveal that current reasoning models do not consistently outperform non-reasoning counterparts on ToM tasks; excessive reasoning can even degrade performance, whereas moderately constraining reasoning length improves results. Moreover, models frequently resort to shortcut strategies based on option matching, and their performance markedly improves when answer choices are removed. To address these issues, the work proposes Slow-to-Fast adaptive reasoning and Think-to-Match anti-shortcut interventions, highlighting the limitations of “slow thinking” in social reasoning and underscoring the need for adaptive reasoning mechanisms.

Technology Category

• Cognitive Modeling & Cognitive Systems: Conceptual Inference and Reasoning
• Machine Learning: Large Multimodal Models (LMMs)
• Knowledge Representation and Reasoning: Reasoning with Beliefs

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Social Networks and Social Media: Generative AI / large language models and their impact on social systems

Abstract

Theory of Mind (ToM) assesses whether models can infer hidden mental states such as beliefs, desires, and intentions, which is essential for natural social interaction. Although recent progress in Large Reasoning Models (LRMs) has boosted step-by-step inference in mathematics and coding, it is still underexplored whether this benefit transfers to socio-cognitive skills. We present a systematic study of nine advanced Large Language Models (LLMs), comparing reasoning models with non-reasoning models on three representative ToM benchmarks. The results show that reasoning models do not consistently outperform non-reasoning models and sometimes perform worse. A fine-grained analysis reveals three insights. First, slow thinking collapses: accuracy significantly drops as responses grow longer, and larger reasoning budgets hurt performance. Second, moderate and adaptive reasoning benefits performance: constraining reasoning length mitigates failure, while distinct success patterns demonstrate the necessity of dynamic adaptation. Third, option matching shortcut: when multiple choice options are removed, reasoning models improve markedly, indicating reliance on option matching rather than genuine deduction. We also design two intervention approaches: Slow-to-Fast (S2F) adaptive reasoning and Think-to-Match (T2M) shortcut prevention to further verify and mitigate the problems. With all results, our study highlights the advancement of LRMs in formal reasoning (e.g., math, code) cannot be fully transferred to ToM, a typical task in social reasoning. We conclude that achieving robust ToM requires developing unique capabilities beyond existing reasoning methods.

Problem

• Theory of Mind
• Large Reasoning Models
• social reasoning
• mental state inference
• reasoning robustness

Innovation

• Theory of Mind
• Large Reasoning Models
• adaptive reasoning
• reasoning shortcuts
• social cognition