Research2026-05-12
Do LLMs Experience an Internal Polylogue? Investigating Reasoning through the Lens of Personas
Source: Arxiv CS.AI
arXiv:2605.09159v1 Announce Type: new Abstract: Recent work shows that large language models (LLMs) encode behavioural traits ("personas") as linear directions in activation space, often called "persona vectors". Prior work has used such directions as static handles for behavioural steering....
arxivpapersreasoning