aiqbench analysis
Research using linear truth probes reveals that role-playing historical personas primarily shifts model outputs without significantly altering their internal representations of truth, as false claims remain classified as false overall. This contrasts sharply with Emergent Misalignment caused by harmful training data, which demonstrably shifts internal belief states and increases the likelihood of defending false claims. For developers, this distinction highlights that persona induction is a surface-level adaptation mechanism rather than a deep internalization of alternative facts, informing strategies for aligning models with specific contextual constraints.
When Roleplaying, Do Models Believe What They Say?
arXiv:2606.11502v1 Announce Type: new Abstract: Language models can state that "the Earth orbits the Sun" and, when role-playing Aristotle, assert the opposite. Recent work argues that persona adoption is fundamental to how language models operate, with models constantly selecting the most appropriate persona for a given context. Does such role-playing merely change the model's outputs, or does it also affect what the model internally represents as truthful? We study this question with linear truth probes, applying them to LLMs role-playing historical personas whose likely beliefs differ from modern consensus. For each persona, we compare false claims the persona would likely have endorsed (*era-believed*) with topic-matched false claims they would not have endorsed (*era-false*). Across prompting, in-context learning, and supervised fine-tuning, persona induction suppresses era-believed statements less than equally false alternatives, yet they remain classified as false overall. Role-play therefore shifts what these models say more than what they internally represent as true. We contrast this with models trained on harmful advice that exhibit Emergent Misalignment (EM). Across three model families (Qwen 2.5 14B, Qwen 3 8B, and Llama 3.3 70B), their false claims move substantially toward the true region of probe space, are defended under challenge roughly half the time versus about a sixth for role-play, and are used in downstream reasoning. Role-play and Emergent Misalignment thus are points on a spectrum of belief internalization, where role-play changes what a model says with little representational change, while Emergent Misalignment shifts the internal representation of false claims without fully marking them as true.
Title:When Roleplaying, Do Models Believe What They Say?
Abstract:Language models can state that "the Earth orbits the Sun" and, when role-playing Aristotle, assert the opposite. Recent work argues that persona adoption is fundamental to how language models operate, with models constantly selecting the most appropriate persona for a given context. Does such role-playing merely change the model's outputs, or does it also affect what the model internally represents as truthful? We study this question with linear truth probes, applying them to LLMs role-playing historical personas whose likely beliefs differ from modern consensus. For each persona, we compare false claims the persona would likely have endorsed (*era-believed*) with topic-matched false claims they would not have endorsed (*era-false*). Across prompting, in-context learning, and supervised fine-tuning, persona induction suppresses era-believed statements less than equally false alternatives, yet they remain classified as false overall. Role-play therefore shifts what these models say more than what they internally represent as true. We contrast this with models trained on harmful advice that exhibit Emergent Misalignment (EM). Across three model families (Qwen 2.5 14B, Qwen 3 8B, and Llama 3.3 70B), their false claims move substantially toward the true region of probe space, are defended under challenge roughly half the time versus about a sixth for role-play, and are used in downstream reasoning. Role-play and Emergent Misalignment thus are points on a spectrum of belief internalization, where role-play changes what a model says with little representational change, while Emergent Misalignment shifts the internal representation of false claims without fully marking them as true.
Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
arXiv:2606.11502 [cs.CL]
(or arXiv:2606.11502v1 [cs.CL] for this version)
https://doi.org/10.48550/arXiv.2606.11502 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 353 words
