# Observer, Not Player: Simulating Theory of Mind in LLMs through Game Observation

arXiv: https://arxiv.org/abs/2512.19210 · HTML: https://mlsanity.com/paper/2512.19210/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 25 of its 44 references that are in the map.
- First posted on arXiv 22 Dec 2025.

## Builds on (most cited first, 25 of 25)

- [Chain-of-Thought Prompting Elicits Reasoning in Large Language Models](https://mlsanity.com/paper/2201.11903/) (2022)
- [Large Language Models are Zero-Shot Reasoners](https://mlsanity.com/paper/2205.11916/) (2022)
- [Training Large Language Models to Reason in a Continuous Latent Space](https://mlsanity.com/paper/2412.06769/) (2024)
- [On the Utility of Learning about Humans for Human-AI Coordination](https://mlsanity.com/paper/1910.05789/) (2019)
- [TextWorld: A Learning Environment for Text-based Games](https://mlsanity.com/paper/1806.11532/) (2018)
- [Large Language Models Fail on Trivial Alterations to Theory-of-Mind Tasks](https://mlsanity.com/paper/2302.08399/) (2023)
- [BabyAI: A Platform to Study the Sample Efficiency of Grounded Language Learning](https://mlsanity.com/paper/1810.08272/) (2018)
- [Unity: A General Platform for Intelligent Agents](https://mlsanity.com/paper/1809.02627/) (2018)
- [Evaluating Large Language Models in Theory of Mind Tasks](https://mlsanity.com/paper/2302.02083/) (2023)
- [Playing repeated games with Large Language Models](https://mlsanity.com/paper/2305.16867/) (2023)
- [The What-If Tool: Interactive Probing of Machine Learning Models](https://mlsanity.com/paper/1907.04135/) (2019)
- [The NetHack Learning Environment](https://mlsanity.com/paper/2006.13760/) (2020)
- [The Hanabi Challenge: A New Frontier for AI Research](https://mlsanity.com/paper/1902.00506/) (2019)
- [OpenSpiel: A Framework for Reinforcement Learning in Games](https://mlsanity.com/paper/1908.09453/) (2019)
- [SmartPlay: A Benchmark for LLMs as Intelligent Agents](https://mlsanity.com/paper/2310.01557/) (2023)
- [GameBench: Evaluating Strategic Reasoning Abilities of LLM Agents](https://mlsanity.com/paper/2406.06613/) (2024)
- [ToMBench: Benchmarking Theory of Mind in Large Language Models](https://mlsanity.com/paper/2402.15052/) (2024)
- [lmgame-Bench: How Good are LLMs at Playing Games?](https://mlsanity.com/paper/2505.15146/) (2025)
- [Theory of Mind in Large Language Models: Examining Performance of 11 State-of-the-Art models vs. Children Aged 7-10 on Advanced Tests](https://mlsanity.com/paper/2310.20320/) (2023)
- [RLCard: A Toolkit for Reinforcement Learning in Card Games](https://mlsanity.com/paper/1910.04376/) (2019)
- [Position: Theory of Mind Benchmarks are Broken for Large Language Models](https://mlsanity.com/paper/2412.19726/) (2024)
- [Evaluating Large Language Models with Grid-Based Game Competitions: An Extensible LLM Benchmark and Leaderboard](https://mlsanity.com/paper/2407.07796/) (2024)
- [Interactive Visualization for Debugging RL](https://mlsanity.com/paper/2008.07331/) (2020)
- [From Code to Play: Benchmarking Program Search for Games Using Large Language Models](https://mlsanity.com/paper/2412.04057/) (2024)
- [Instruction-Driven Game Engines on Large Language Models](https://mlsanity.com/paper/2404.00276/) (2024)

## Abstract (arXiv)

We present an interactive framework for evaluating whether large language models (LLMs) exhibit genuine "understanding" in a simple yet strategic environment. As a running example, we focus on Rock-Paper-Scissors (RPS), which, despite its apparent simplicity, requires sequential reasoning, adaptation, and strategy recognition. Our system positions the LLM as an Observer whose task is to identify which strategies are being played and to articulate the reasoning behind this judgment. The purpose is not to test knowledge of Rock-Paper-Scissors itself, but to probe whether the model can exhibit mind-like reasoning about sequential behavior. To support systematic evaluation, we provide a benchmark consisting of both static strategies and lightweight dynamic strategies specified by well-prompted rules. We quantify alignment between the Observer's predictions and the ground-truth distributions induced by actual strategy pairs using three complementary signals: Cross-Entropy, Brier score, and Expected Value (EV) discrepancy. These metrics are further integrated into a unified score, the Union Loss, which balances calibration, sensitivity, and payoff alignment. Together with a Strategy Identification Rate (SIR) metric, our framework captures not only predictive accuracy but also whether the model can stably identify the latent strategies in play. The demo emphasizes interactivity, transparency, and reproducibility. Users can adjust LLM distributions in real time, visualize losses as they evolve, and directly inspect reasoning snippets to identify where and why failures occur. In doing so, our system provides a practical and interpretable proxy for mind-like inference in sequential games, offering insights into both the strengths and limitations of current LLM reasoning.