{"slug": "role-steering-of-language-models-for-social-simulations", "title": "Role Steering of Language Models for Social Simulations", "summary": "Researchers introduced an activation-steering screening workflow for role-conditioned language-model agents, testing it on OLMo-3-7B-Instruct with a 275-role inventory and 228 role-agnostic questions. Role-specific directions achieved higher judged role-profile alignment (mean 63.2) than an assistant-axis control (41.1), but 38 roles declined across all six measured dimensions, indicating the need for per-role coefficient selection. The code and evaluation artifacts are available at https://anonymous.4open.science/r/anonymous-research-code-5F03/.", "body_md": "arXiv:2608.00023v1 Announce Type: new\nAbstract: Social simulations built from language-model agents need role-conditioned behavior that can be checked before agents are placed into a simulated population. We introduce an activation-steering screening workflow for role-conditioned agents: define a role profile, extract a role-specific direction, sweep four steering coefficients, evaluate role-profile alignment, and pass or flag each candidate configuration. On OLMo-3-7B-Instruct, we apply the workflow to a mixed 275-role inventory with 228 role-agnostic questions, GPT-4.1-mini prompted role references, and GPT-4.1-mini judges. Role-specific directions receive higher judged role-profile alignment than an assistant-axis directional control from prior persona-vector work, with mean overall scores of 63.2 versus 41.1 across the tested grid. They also preserve high lexical diversity, while the control drops sharply at larger coefficients. The role-level screen is the main practical output: most roles improve as steering increases, but 38 roles decline across all six measured dimensions, showing why simulation builders should choose coefficients per role rather than deploy a uniform high-strength setting. We make our code and evaluation artifacts available at https://anonymous.4open.science/r/anonymous-research-code-5F03/.", "url": "https://wpnews.pro/news/role-steering-of-language-models-for-social-simulations", "canonical_source": "https://arxiv.org/abs/2608.00023", "published_at": "2026-08-04 04:00:00+00:00", "updated_at": "2026-08-04 04:35:45.661816+00:00", "lang": "en", "topics": ["large-language-models", "ai-research"], "entities": ["OLMo-3-7B-Instruct", "GPT-4.1-mini"], "alternates": {"html": "https://wpnews.pro/news/role-steering-of-language-models-for-social-simulations", "markdown": "https://wpnews.pro/news/role-steering-of-language-models-for-social-simulations.md", "text": "https://wpnews.pro/news/role-steering-of-language-models-for-social-simulations.txt", "jsonld": "https://wpnews.pro/news/role-steering-of-language-models-for-social-simulations.jsonld"}}