Theory of Mind in Large Language Models: Assessment and Enhancement (2025.acl-long)
Copied to clipboard
| Challenge: | Theory of Mind (ToM) is a cornerstone of human social intelligence . Large Language Models (LLMs) are increasingly integrated into daily life . |
| Approach: | They analyze evaluation benchmarks and enhancement strategies to evaluate LLMs' ToM capabilities. |
| Outcome: | The proposed and widely used story-based benchmarks and enhancement strategies are used to evaluate LLMs' ToM capabilities. |
Similar Papers
Agentic-ToM: Cognition-Inspired Agentic Processing For Enhancing Theory of Mind Reasoning (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Current models struggle with reasoning about others’ perspectives, limiting their ability to attribute mental states to oneself and others. |
| Approach: | They propose to embed psychologically-grounded functions into LLMs to enable them to attribute mental states to oneself and others, known as Theory of Mind. |
| Outcome: | The proposed approach outperforms baselines on three ToM datasets without task-specific modifications. |
Does Theory of Mind Improvement Really Benefit Human-AI Interactions? Empirical Findings from Interactive Evaluations (2026.findings-acl)
Copied to clipboard
| Challenge: | Existing benchmarks measure ToM capability improvement through story-reading, multiple-choice questions from a third-person perspective, while ignoring the first-person, dynamic nature of human-AI interactions. |
| Approach: | They propose a new paradigm of interactive ToM evaluation with both perspective and metric shifts. |
| Outcome: | The proposed approach improves the performance of four representative LLM enhancement techniques using real-world datasets and a user study. |
Hi-ToM: A Benchmark for Evaluating Higher-Order Theory of Mind Reasoning in Large Language Models (2023.findings-emnlp)
Copied to clipboard
| Challenge: | Theory of Mind (ToM) is the ability to reason about one's own and others' mental states. |
| Approach: | They propose a higher-order theory of mind benchmark and introduce a new deception mechanism to evaluate ToM reasoning. |
| Outcome: | The proposed benchmarks show that the LLMs are not performing well on higher-order tasks. |
XToM: Exploring the Multilingual Theory of Mind for Large Language Models (2026.acl-long)
Copied to clipboard
Chunkit Chan, Yauwai Yim, Hongchuan Zeng, Zhiying Zou, Xinyuan Cheng, Zhifan Sun, Zheye Deng, Kawai Chung, Yuzhuo Ao, Fan Yixiang, Cheng Jiayang, Ercong Nie, Ginny Wong, Helmut Schmid, Hinrich Schuetze, Simon See, Yangqiu Song
| Challenge: | Existing evaluations of ToM in LLMs are limited to English, neglecting the linguistic diversity that shapes human cognition. |
| Approach: | They propose a multilingual benchmark that evaluates ToM across five languages . they find that models excel in multilingual language understanding, but their ToM performance varies across languages. |
| Outcome: | The proposed benchmark evaluates LLMs across five languages and incorporates diverse task scenarios. |
Decompose-ToM: Enhancing Theory of Mind Reasoning in Large Language Models through Simulation and Task Decomposition (2025.coling-main)
Copied to clipboard
| Challenge: | Theory of Mind (ToM) is the ability to attribute and infer the mental states of others. |
| Approach: | They propose an LLM-based inference algorithm that improves model performance on complex ToM tasks by simulating user perspectives. |
| Outcome: | The proposed algorithm improves model performance on complex ToM tasks while requiring minimal prompt tuning across tasks and no additional model training. |
Views Are My Own, but Also Yours: Benchmarking Theory of Mind Using Common Ground (2024.findings-acl)
Copied to clipboard
Adil Soubki, John Murzaku, Arash Yousefi Jordehi, Peter Zeng, Magdalena Markowska, Seyed Abolghasem Mirroshandel, Owen Rambow
| Challenge: | Existing benchmarks for theory of mind (ToM) use synthetic data, which can misalign with human behavior. |
| Approach: | They propose a question-answer benchmark based on naturally occurring spoken dialogs to evaluate theory of mind capabilities of language models. |
| Outcome: | The proposed dataset shows that LMs struggle to demonstrate theory of mind (ToM) . |
MindGames: Targeting Theory of Mind in Large Language Models with Dynamic Epistemic Modal Logic (2023.findings-emnlp)
Copied to clipboard
| Challenge: | Theory of Mind (ToM) is a critical component of intelligence but its assessment remains the subject of heated debates. |
| Approach: | They propose to use dynamic epistemic logic to isolate a particular component of ToM and generate controlled problems in English natural language. |
| Outcome: | The proposed language model scales from 70M to 6B and 350M to 174B do not consistently yield better results than random chance. |
Clever Hans or Neural Theory of Mind? Stress Testing Social Reasoning in Large Language Models (2024.eacl-long)
Copied to clipboard
Natalie Shapira, Mosh Levy, Seyed Hossein Alavi, Xuhui Zhou, Yejin Choi, Yoav Goldberg, Maarten Sap, Vered Shwartz
| Challenge: | Recent work suggests that Large Language Models (LLMs) exhibit Neural Theory-of-Mind (N-ToM) however, prior work reached conflicting conclusions regarding those abilities. |
| Approach: | They examine the extent of Large Language Models’ N-ToM abilities through an extensive evaluation of 6 tasks and find that LLMs struggle with adversarial examples . |
| Outcome: | The proposed metrics show that LLMs exhibit certain N-ToM abilities, but this behavior is far from robust. |
Beyond Words: Integrating Theory of Mind into Conversational Agents for Human-Like Belief, Desire, and Intention Alignment (2025.findings-acl)
Copied to clipboard
| Challenge: | Empirical evaluations of LLaMA-3 models demonstrate that ToM-informed alignment improves response quality, achieving win rates of 63% and 67%, respectively. |
| Approach: | They investigate whether open-source LLaMA models can represent and retain ToM-related constructs and whether they can be used to generate more aligned responses. |
| Outcome: | The proposed models can represent and retain ToM-related constructs and improve response quality. |
OpenToM: A Comprehensive Benchmark for Evaluating Theory-of-Mind Reasoning Capabilities of Large Language Models (2024.acl-long)
Copied to clipboard
| Challenge: | Existing N-ToM benchmarks lack ambiguous and artificial narratives, lack of personality traits and preferences, and limited diversity in the questions posed. |
| Approach: | They propose a benchmark to assess Neural Theory-of-Mind (N-ToM) with longer and clearer narrative stories, characters with explicit personality traits, actions triggered by character intentions, and questions designed to challenge LLMs’ abilities of modeling characters’ mental states. |
| Outcome: | The proposed test aims to assess the performance of LLMs in the physical and psychological worlds. |