Pith. sign in

REVIEW 5 cited by

Think Twice: Perspective-Taking Improves Large Language Models' Theory-of-Mind Capabilities

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2311.10227 v1 pith:Q56VHHRF submitted 2023-11-16 cs.AI cs.CL

Think Twice: Perspective-Taking Improves Large Language Models' Theory-of-Mind Capabilities

classification cs.AI cs.CL
keywords capabilitiesperspective-takingtheoryllmscognitivelanguagelargemental
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
read the original abstract

Human interactions are deeply rooted in the interplay of thoughts, beliefs, and desires made possible by Theory of Mind (ToM): our cognitive ability to understand the mental states of ourselves and others. Although ToM may come naturally to us, emulating it presents a challenge to even the most advanced Large Language Models (LLMs). Recent improvements to LLMs' reasoning capabilities from simple yet effective prompting techniques such as Chain-of-Thought have seen limited applicability to ToM. In this paper, we turn to the prominent cognitive science theory "Simulation Theory" to bridge this gap. We introduce SimToM, a novel two-stage prompting framework inspired by Simulation Theory's notion of perspective-taking. To implement this idea on current ToM benchmarks, SimToM first filters context based on what the character in question knows before answering a question about their mental state. Our approach, which requires no additional training and minimal prompt-tuning, shows substantial improvement over existing methods, and our analysis reveals the importance of perspective-taking to Theory-of-Mind capabilities. Our findings suggest perspective-taking as a promising direction for future research into improving LLMs' ToM capabilities.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 5 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. DialToM: A Theory of Mind Benchmark for Forecasting State-Driven Dialogue Trajectories

    cs.CL 2026-04 conditional novelty 6.5

    LLMs excel at retrospective mental-state labeling on naturalistic dialogues but mostly fail a context-free prospective probe that maps isolated mental-state profiles to dialogue trajectories, despite expert 100% accuracy.

  2. DialToM: A Theory of Mind Benchmark for Forecasting State-Driven Dialogue Trajectories

    cs.CL 2026-04 unverdicted novelty 6.0

    LLMs identify mental states in dialogues well but mostly fail to forecast state-consistent future trajectories, except Gemini 3 Pro, with only weak overlap to human inferences.

  3. From Fallback to Frontline: When Can LLMs be Superior Annotators of Human Perspectives?

    cs.AI 2026-04 unverdicted novelty 6.0

    LLMs can be statistically superior to humans at estimating group-level judgments on subjective tasks because of their low variance and decoupled representation-processing biases.

  4. MeTHanol: Modularized Thinking Language Models with Intermediate Layer Thinking, Decoding and Bootstrapping Reasoning

    cs.CL 2024-09 unverdicted novelty 6.0

    MeTHanol fine-tunes an intermediate LLM layer to generate thoughts in a first pass, then uses those thoughts for a second-pass answer, showing gains on Theory of Mind and vignette tasks plus adaptation to character prompts.

  5. When Researchers Say Mental Model/Theory of Mind of AI, What Are They Really Talking About?

    cs.HC 2025-10 conditional novelty 3.0

    Researchers' claims of AI theory of mind are really about behavioral prediction, so AI evaluation should shift from isolated cognitive tests to human-AI interaction.