mlsanityData to 14 Oct 2026

GroupToM-Bench: Benchmarking Group Theory of Mind and Nonlinear Social Emergence in MLLMs

arXiv 2606.04184 ↗First posted 2 Jun 2026cs.CVcited by 1

On the map: World models → World models for language agents

Abstract (arXiv)

True general intelligence requires not only a model of the physical world but also a social world model: the capacity to infer how individual mental states interact and crystallize into group-level outcomes. Despite notable progress in individual-level Theory of Mind (ToM) reasoning, existing multimodal large language models fail at this broader task. Collective behavior emerges non-linearly from social tensions, conformity dynamics, and structural constraints, meaning it cannot be recovered by merely summing individual intentions. We present GroupToM-Bench, the first multimodal benchmark for group-level ToM, built around a causal chain spanning micro-level BDI states (belief, desire, intention), meso-level group tension and structural constraints, and macro-level outcome prediction and mechanistic attribution. To probe this full arc, we develop a seven-level cognitive audit framework. Experiments reveal a gap between current models and human baselines, highlighting a failure to process social structures and non-linear collective dynamics.

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Placed in World models for language agents, one of 11 directions in World models. Builds on 14 of its 54 references that are in the map. First posted on arXiv 2 Jun 2026.

Builds on

14 of 54 references.

References · 14

  1. 2018World Modelscited by 1,513
  2. 2023MM-Vet: Evaluating Large Multimodal Models for Integrated Capabilitiescited by 1,239
  3. 2025More Thinking, Less Seeing? Assessing Amplified Hallucination in Multimodal Reasoning Modelscited by 86
  4. 2024SimpleToM: Exposing the Gap between Explicit ToM Inference and Implicit ToM Application in LLMscited by 57
  5. 2024NegotiationToM: A Benchmark for Stress-testing Machine Theory of Mind on Negotiation Surroundingcited by 51
  6. 2025SoMi-ToM: Evaluating Multi-Perspective Theory of Mind in Embodied Social Interactionscited by 16
  7. 2025Rethinking Theory of Mind Benchmarks for LLMs: Towards A User-Centered Perspectivecited by 14
  8. 2025MOMENTS: A Comprehensive Multimodal Benchmark for Theory of Mindcited by 13
  9. 2025EgoToM: Benchmarking Theory of Mind Reasoning from Egocentric Videoscited by 12
  10. 2025SocialMaze: A Benchmark for Evaluating and Enhancing Social Reasoning in Large Language Models in Complex Social Environmentscited by 11
  11. 2025Can MLLMs Read the Room? A Multimodal Benchmark for Verifying Truthfulness in Multi-Party Social Interactionscited by 6
  12. 2024Diversifying Policy Behaviors with Extrinsic Behavioral Curiositycited by 4
2 more
  1. 2025The Decrypto Benchmark for Multi-Agent Reasoning and Theory of Mindcited by 4
  2. 2026CaveAgent: Transforming LLMs into Stateful Runtime Operatorscited by 3

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026Unified Hallucination Fuzzing for Multimodal Large Language Modelscited by 1

Similar, not cited

Close in wording, no citation either way

  1. 2026MeetingToM: Evaluating Multimodal LLMs on Theory-of-Mind Reasoning in Multi-Party Meetings
  2. 2023Understanding Social Reasoning in Language Models with Language Models
  3. 2026CoMMET: A Psychologically Grounded Benchmark for Evaluating Theory of Mind in Multimodal LLMs
  4. 2026Belief Without Behavior: Measuring the Translation of Theory of Mind into Coordinated Social Action in Vision-Language Models
  5. 2025Do Theory of Mind Benchmarks Need Explicit Human-like Reasoning in Language Models?
  6. 2024Multi-ToM: Evaluating Multilingual Theory of Mind Capabilities in Large Language Models
  7. 2024Language Models Represent Beliefs of Self and Others
  8. 2026On Emergent Social World Models -- Evidence for Functional Integration of Theory of Mind and Pragmatic Reasoning in Language Models