arXiv AI July 24, 2026 By Adam Rigby, Raz Saremi, Azadeh Sohrabinejad, Mehdi Rahimi

The Storyteller in the Model: Narrative Pattern Inheritance, Escalation Dynamics, and Alignment Governance in LLMs

arXiv:2607. 20449v1 Announce Type: cross Abstract: LLMs are trained predominantly on human-authored text, yet the structural and narrative conventions embedded in that text are rarely examined as a source of systematic behavioral influence, or as a governance risk in deployed systems.

Summary generated by The Flow from the publisher's feed. The full article lives at arXiv AI.

arXiv AI

Jul 22

The Story Shapes the Agent: Narrative Priors in LLM Behavior

arXiv:2607. 18566v1 Announce Type: cross 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.

By Yixuan Wang, James Lester, Shashank Srivastava

llms agents

Hugging Face Trending Papers

Aug 11

Data Attribution of Emergent Misalignment with Persona Features

Emergent misalignment (EM) is the phenomenon where fine-tuning a language model on a narrow task leads to harmful behavior in unrelated domains. A leading mechanistic account attributes EM to persona features: latent directions acquired during pre-training that misaligned fine-tuning amplifies.

llms robotics fine-tuning safety

arXiv AI

Jul 28

Do LLMs Experience an Internal Polylogue? Investigating Reasoning through the Lens of Personas

arXiv:2605. 09159v2 Announce Type: replace Abstract: Recent work shows that large language models (LLMs) encode behavioral traits ("personas") as linear directions in activation space, often called "persona vectors".

By Nils A. Herrmann, Leander Girrbach, Kirill Bykov, Zeynep Akata

llms

arXiv AI

Jul 20

The AI Fiction Paradox

arXiv:2603. 13545v2 Announce Type: replace Abstract: AI development has a fiction dependency problem.

By Katherine Elkins

agents robotics

Hugging Face Trending Papers

Jun 17

The Wrong Kind of Right: Quantifying and Localizing Misfired Alignment in LLMs

Warning: This paper studies stereotypes and biases, and contains potentially disturbing examples, used for illustration purposes only. Our findings should not be interpreted as an argument against alignment.

llms benchmarks safety

arXiv AI

Jun 6

How Far Did They Go? The Persuasive Tactics of Covert LLM Agents in a Discontinued Field Experiment

arXiv:2606. 05256v1 Announce Type: new Abstract: This study analyzes a publicly released dataset from a discontinued field experiment on Reddit's r/ChangeMyView.

By Kokil Jaidka, Saifuddin Ahmed

llms agents safety

Related stories

The Story Shapes the Agent: Narrative Priors in LLM Behavior

Data Attribution of Emergent Misalignment with Persona Features

Do LLMs Experience an Internal Polylogue? Investigating Reasoning through the Lens of Personas

The AI Fiction Paradox

The Wrong Kind of Right: Quantifying and Localizing Misfired Alignment in LLMs

How Far Did They Go? The Persuasive Tactics of Covert LLM Agents in a Discontinued Field Experiment