The Story Shapes the Agent: Narrative Priors in LLM Behavior

• URL: http://arxiv.org/abs/2607.18566v1
• Date: Mon, 20 Jul 2026 23:03:16 GMT
• Title: The Story Shapes the Agent: Narrative Priors in LLM Behavior
• Abstract summary: Persona prompting is widely used to steer LLM agent behavior, yet the narrative framing of a task can matter more than the assigned persona.<n>We construct three text-based investigation games that share the same action space, stage progression, and resource constraints while varying only task narrative.<n>We identify narrative priors: systematic action tendencies activated by a task's story framing, independent of its decision structure.
• Score: 12.905014436867317
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Persona prompting is widely used to steer LLM agent behavior, yet the narrative framing of a task can matter more than the assigned persona. We isolate this effect through structural isomorphism, constructing three text-based investigation games that share the same action space, stage progression, and resource constraints while varying only task narrative: disease investigation, IT troubleshooting, and murder mystery. Across 1,890 sessions spanning 3 models and 10 personas, we identify narrative priors: systematic action tendencies activated by a task's story framing, independent of its decision structure. Narrative priors explain 5-31x more behavioral variance than persona, are consistent across model architectures, and in two of three domains are negatively associated with task success. Persona effects that do transfer across narratives arise from behavioral anchors, persona descriptions whose language maps directly onto shared actions. Causal interventions confirm this: removing anchor words from a high-transfer persona reduces cross-narrative consistency by 95%. Our framework also generalizes to a held-out fourth narrative and yields a persona-selection method that improves cross-narrative transfer. These results suggest that LLM behavior that survives narrative changes should be grounded in concrete actions rather than abstract descriptions.

Related papers

• Bad company corrupts good morals: Understanding and Measuring Narrative-Induced Moral Reasoning Degradation in LLMs [9.12926114428835]
  Large language models are deployed in long-context, emotionally interactive environments like digital humans, AI companions, educational assistants, and counseling systems.<n>This raises an important question: can prolonged narrative exposure reshape the reasoning and alignment stability of LLMs?<n>We present the first systematic study of narrative-induced alignment degradation in LLMs.
  arXiv  Detail & Related papers  (2026-06-27T15:30:42Z)
• Characterizing Narrative Content in Web-scale LLM Pretraining Data [14.02632268506113]
  We present the first fine-grained study of narrative features in Dolma, a 3-trillion-token open pretraining corpus.<n>We design a framework spanning three core narrative elements (agency, setting, and events) operationalized as 11 interpretable dimensions.<n>We apply NarraBERT to 3M passages, resulting in a new dataset, NarraDolma.
  arXiv  Detail & Related papers  (2026-06-17T18:03:34Z)
• From We to Me: Theory Informed Narrative Shift with Abductive Reasoning [1.0930816518774933]
  We propose a neurosymbolic approach grounded in social science theory and abductive reasoning.<n>Our method automatically extracts rules to abduce the specific story elements needed to guide an Large Language Model.<n>We show similar performance across both directions for Llama-4, and Grok-4 and competitive performance for Deep-seek-R1.
  arXiv  Detail & Related papers  (2026-02-10T04:10:19Z)
• NarraScore: Bridging Visual Narrative and Musical Dynamics via Hierarchical Affective Control [59.6128550986024]
  NarraScore is a hierarchical framework predicated on the core insight that emotion serves as a high-density compression of narrative logic.<n>NarraScore employs a Dual-Branch Injection strategy to reconcile global structure with local dynamism.<n>NarraScore achieves state-of-the-art consistency and narrative alignment with negligible computational overhead.
  arXiv  Detail & Related papers  (2026-02-09T09:39:42Z)
• Contrastive Learning with Narrative Twins for Modeling Story Salience [19.936257724469783]
  We present a contrastive learning framework for modeling narrative salience.<n>Our model is trained to distinguish a story from both its narrative twin and a distractor with similar surface features but different plot.<n>Experiments on short narratives from the ROCStories corpus and longer Wikipedia plot summaries show that contrast learned story embeddings outperform a masked-language-model baseline.
  arXiv  Detail & Related papers  (2026-01-12T17:48:46Z)
• Living the Novel: A System for Generating Self-Training Timeline-Aware Conversational Agents from Novels [50.43968216132018]
  We present an end-to-end system that transforms any literary work into an immersive, multi-character conversational experience.<n>This system is designed to solve two fundamental challenges for LLM-driven characters.
  arXiv  Detail & Related papers  (2025-12-08T11:57:46Z)
• Failure Modes of LLMs for Causal Reasoning on Narratives [51.19592551510628]
  We investigate the interaction between world knowledge and logical reasoning.<n>We find that state-of-the-art large language models (LLMs) often rely on superficial generalizations.<n>We show that simple reformulations of the task can elicit more robust reasoning behavior.
  arXiv  Detail & Related papers  (2024-10-31T12:48:58Z)
• Causal Micro-Narratives [62.47217054314046]
  We present a novel approach to classify causal micro-narratives from text. These narratives are sentence-level explanations of the cause(s) and/or effect(s) of a target subject.
  arXiv  Detail & Related papers  (2024-10-07T17:55:10Z)
• Mapping News Narratives Using LLMs and Narrative-Structured Text Embeddings [0.0]
  We introduce a numerical narrative representation grounded in structuralist linguistic theory. We extract the actants using an open-source LLM and integrate them into a Narrative-Structured Text Embedding. We demonstrate the analytical insights of the method on the example of 5000 full-text news articles from Al Jazeera and The Washington Post on the Israel-Palestine conflict.
  arXiv  Detail & Related papers  (2024-09-10T14:15:30Z)
• Are Large Language Models Capable of Generating Human-Level Narratives? [114.34140090869175]
  This paper investigates the capability of LLMs in storytelling, focusing on narrative development and plot progression. We introduce a novel computational framework to analyze narratives through three discourse-level aspects. We show that explicit integration of discourse features can enhance storytelling, as is demonstrated by over 40% improvement in neural storytelling.
  arXiv  Detail & Related papers  (2024-07-18T08:02:49Z)
• TEACH: Temporal Action Composition for 3D Humans [50.97135662063117]
  Given a series of natural language descriptions, our task is to generate 3D human motions that correspond semantically to the text. In particular, our goal is to enable the synthesis of a series of actions, which we refer to as temporal action composition.
  arXiv  Detail & Related papers  (2022-09-09T00:33:40Z)

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.