Language Models Might Not Understand You: Evaluating Theory of Mind via Story Prompting

• URL: http://arxiv.org/abs/2506.19089v2
• Date: Mon, 30 Jun 2025 19:05:37 GMT
• Title: Language Models Might Not Understand You: Evaluating Theory of Mind via Story Prompting
• Abstract summary: $textttStorySim$ is a programmable framework for synthetically generating stories.<n>It produces novel, compositional story prompts anchored by a highly controllable $textttStoryboard$.<n>Our framework enabled us to find evidence of recency bias and an over-reliance on earlier events in the story.
• Score: 4.373803477995854
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: We introduce $\texttt{StorySim}$, a programmable framework for synthetically generating stories to evaluate the theory of mind (ToM) and world modeling (WM) capabilities of large language models (LLMs). Unlike prior benchmarks that may suffer from contamination in pretraining data, $\texttt{StorySim}$ produces novel, compositional story prompts anchored by a highly controllable $\texttt{Storyboard}$, enabling precise manipulation of character perspectives and events. We use this framework to design first- and second-order ToM tasks alongside WM tasks that control for the ability to track and model mental states. Our experiments across a suite of state-of-the-art LLMs reveal that most models perform better on WM tasks than ToM tasks, and that models tend to perform better reasoning with humans compared to inanimate objects. Additionally, our framework enabled us to find evidence of heuristic behavior such as recency bias and an over-reliance on earlier events in the story. All code for generating data and evaluations is freely available.

Related papers

• StoryAlign: Evaluating and Training Reward Models for Story Generation [52.44891270363296]
  We introduce StoryRMB, the first benchmark for assessing reward models on story preferences.<n>We develop StoryReward, an advanced reward model for story preference trained on this dataset.<n>StoryReward achieves state-of-the-art (SoTA) performance on StoryRMB, outperforming much larger models.
  arXiv  Detail & Related papers  (2026-05-06T12:28:17Z)
• Decompose-ToM: Enhancing Theory of Mind Reasoning in Large Language Models through Simulation and Task Decomposition [2.089191490381739]
  Theory of Mind (ToM) is the ability to understand and reflect on the mental states of others.<n>Large Language Models (LLMs) possess only a rudimentary understanding of ToM.<n>We propose Decompose-ToM'': an LLM-based inference algorithm that improves model performance on complex ToM tasks.
  arXiv  Detail & Related papers  (2025-01-15T18:44:01Z)
• Explore Theory of Mind: Program-guided adversarial data generation for theory of mind reasoning [88.68573198200698]
  We introduce ExploreToM, the first framework to allow large-scale generation of diverse and challenging theory of mind data.<n>Our approach leverages an A* search over a custom domain-specific language to produce complex story structures and novel, diverse, yet plausible scenarios.<n>Our evaluation reveals that state-of-the-art LLMs, such as Llama-3.1-70B and GPT-4o, show accuracies as low as 0% and 9% on ExploreToM-generated data.
  arXiv  Detail & Related papers  (2024-12-12T21:29:00Z)
• Not (yet) the whole story: Evaluating Visual Storytelling Requires More than Measuring Coherence, Grounding, and Repetition [8.058451580903123]
  We introduce a novel method that measures story quality in terms of human likeness. We then use this method to evaluate the stories generated by several models. Upgrading the visual and language components of TAPM results in a model that yields competitive performance.
  arXiv  Detail & Related papers  (2024-07-05T14:48:15Z)
• LangSuitE: Planning, Controlling and Interacting with Large Language Models in Embodied Text Environments [70.91258869156353]
  We introduce LangSuitE, a versatile and simulation-free testbed featuring 6 representative embodied tasks in textual embodied worlds. Compared with previous LLM-based testbeds, LangSuitE offers adaptability to diverse environments without multiple simulation engines. We devise a novel chain-of-thought (CoT) schema, EmMem, which summarizes embodied states w.r.t. history information.
  arXiv  Detail & Related papers  (2024-06-24T03:36:29Z)
• Zero, Finite, and Infinite Belief History of Theory of Mind Reasoning in Large Language Models [5.455744338342196]
  Large Language Models (LLMs) have recently shown a promise and emergence of Theory of Mind (ToM) ability. We propose a novel concept, taxonomy, and framework, the ToM reasoning with Zero, Finite, and Infinite Belief History. We have evaluated six LLMs with this game and found their performance on Zero Belief History is consistently better than on Finite Belief History.
  arXiv  Detail & Related papers  (2024-06-07T10:04:39Z)
• TARN-VIST: Topic Aware Reinforcement Network for Visual Storytelling [14.15543866199545]
  As a cross-modal task, visual storytelling aims to generate a story for an ordered image sequence automatically. We propose a novel method, Topic Aware Reinforcement Network for VIsual StoryTelling (TARN-VIST) In particular, we pre-extracted the topic information of stories from both visual and linguistic perspectives.
  arXiv  Detail & Related papers  (2024-03-18T08:01:23Z)
• Language-Guided World Models: A Model-Based Approach to AI Control [31.9089380929602]
  This paper introduces the concept of Language-Guided World Models (LWMs) LWMs are probabilistic models that can simulate environments by reading texts. We take initial steps in developing robust LWMs that can generalize to compositionally novel language descriptions.
  arXiv  Detail & Related papers  (2024-01-24T03:11:36Z)
• MMToM-QA: Multimodal Theory of Mind Question Answering [80.87550820953236]
  Theory of Mind (ToM) is an essential ingredient for developing machines with human-level social intelligence. Recent machine learning models, particularly large language models, seem to show some aspects of ToM understanding. Human ToM, on the other hand, is more than video or text understanding. People can flexibly reason about another person's mind based on conceptual representations extracted from any available data.
  arXiv  Detail & Related papers  (2024-01-16T18:59:24Z)
• CLOMO: Counterfactual Logical Modification with Large Language Models [109.60793869938534]
  We introduce a novel task, Counterfactual Logical Modification (CLOMO), and a high-quality human-annotated benchmark. In this task, LLMs must adeptly alter a given argumentative text to uphold a predetermined logical relationship. We propose an innovative evaluation metric, the Self-Evaluation Score (SES), to directly evaluate the natural language output of LLMs.
  arXiv  Detail & Related papers  (2023-11-29T08:29:54Z)
• Large Language Models with Controllable Working Memory [64.71038763708161]
  Large language models (LLMs) have led to a series of breakthroughs in natural language processing (NLP) What further sets these models apart is the massive amounts of world knowledge they internalize during pretraining. How the model's world knowledge interacts with the factual information presented in the context remains under explored.
  arXiv  Detail & Related papers  (2022-11-09T18:58:29Z)
• Robust Preference Learning for Storytelling via Contrastive Reinforcement Learning [53.92465205531759]
  Controlled automated story generation seeks to generate natural language stories satisfying constraints from natural language critiques or preferences. We train a contrastive bi-encoder model to align stories with human critiques, building a general purpose preference model. We further fine-tune the contrastive reward model using a prompt-learning technique to increase story generation robustness.
  arXiv  Detail & Related papers  (2022-10-14T13:21:33Z)
• StoryDALL-E: Adapting Pretrained Text-to-Image Transformers for Story Continuation [76.44802273236081]
  We develop a model StoryDALL-E for story continuation, where the generated visual story is conditioned on a source image. We show that our retro-fitting approach outperforms GAN-based models for story continuation and facilitates copying of visual elements from the source image. Overall, our work demonstrates that pretrained text-to-image synthesis models can be adapted for complex and low-resource tasks like story continuation.
  arXiv  Detail & Related papers  (2022-09-13T17:47:39Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.