Research map / Large language models / Agents, code and math
SimpleToM: Exposing the Gap between Explicit ToM Inference and Implicit ToM Application in LLMs
Yuling Gu, Oyvind Tafjord, Hyunwoo Kim, Jared Moore, Ronan Le Bras, Peter Clark, Yejin Choi
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| ask behavior prediction instead of mental state question (applied vs explicit ToM) | GPT-4o · SimpleToM | accuracy (%): 95.6 -> 49.5 | Hurt |
| ask judgment question instead of mental state question | Llama-3.1-405B · SimpleToM | accuracy (%): 97.8 -> 10.0 | Hurt |
| add generic chain-of-thought prompt (ablation) | GPT-4o · SimpleToM behavior prediction | accuracy (%): 49.5 -> 62.8 | Improved |
| add generic chain-of-thought prompt (ablation) | GPT-4o · SimpleToM judgment | accuracy (%): 15.3 -> 39.2 | Improved |
| add mental state reminder to prompt (ablation) | Llama-3.1-405B · SimpleToM behavior prediction | accuracy (%): 58.2 -> 89.5 | Improved |
| add mental state reminder to prompt (ablation) | GPT-4o · SimpleToM judgment | accuracy (%): 15.3 -> 42.2 | Improved |
| add mental state reminder to prompt (ablation) | Claude-3.5-Sonnet · SimpleToM judgment | accuracy (%): 24.9 -> 84.1 | Improved |
| add generic system prompt (SysP) about considering mental states (ablation) | GPT-4o, Llama-3.1-405B, Claude-3.5-Sonnet · SimpleToM behavior prediction | accuracy change: -2.2% to +6.3% | No clear effect |
| add explicit awareness hint to system prompt (SysP*) (ablation) | GPT-4o, Llama-3.1-405B, Claude-3.5-Sonnet · SimpleToM behavior prediction | accuracy change: +19.1% to +25.1% | Improved |
| use awareness-specific CoT* prompt instead of generic CoT (ablation) | GPT-4o · SimpleToM judgment | accuracy (%): 39.2 -> 86.7 | Improved |
| combine mental state reminder with CoT* (ablation) | Claude-3.5-Sonnet · SimpleToM | average accuracy (%): 63.3 -> 97.1 | Improved |
| use reasoning model (o1-preview) instead of non-reasoning models | o1-preview · SimpleToM behavior prediction | accuracy (%): 84.1 | Improved |
Closest papers by meaning
- Theory of Mind and Persuasion Beyond Conversation: Assessing the Capacity of LLMs to Induce Belief States via Planning and Action (NeurIPS 2026 · similarity 0.91)
- Language Models Represent Beliefs of Self and Others (ICML 2024 · similarity 0.91)
- OmniToM: Benchmarking Theory of Mind in LLMs via Explicit Belief Modeling (NeurIPS 2026 · similarity 0.91)
- AutoToM: Scaling Model-based Mental Inference via Automated Agent Modeling (NeurIPS 2025 · similarity 0.90)
- ToMATO: Verbalizing the Mental States of Role-Playing LLMs for Benchmarking Theory of Mind (AAAI 2025 · similarity 0.90)
- MetaMind: Modeling Human Social Thoughts with Metacognitive Multi-Agent Systems (NeurIPS 2025 · similarity 0.89)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
