{"slug": "dymt-esb-dynamic-multi-turn-evaluation-of-social-bias-in-user-llm-interactions", "title": "DyMT-ESB: Dynamic Multi-Turn Evaluation of Social Bias in User-LLM Interactions", "summary": "A new arXiv paper, DyMT-ESB, introduces a controlled evaluation protocol for social bias in multi-turn user-LLM interactions that generates follow-up user queries from evolving dialogue history rather than relying on pre-specified or template-based inputs. The authors report that LLMs exhibit social bias even in coherent, response-conditioned multi-turn interactions, revealing late-emerging bias, non-monotonic bias patterns, and bias re-emergence, findings they say motivate evaluations extending beyond fixed-turn, pre-scripted protocols.", "body_md": "arXiv:2609.18649v1 Announce Type: new \nAbstract: Warning: This paper contains examples of stereotypes and social bias. LLMs are increasingly used in interactive settings by the general public, making the evaluation of model behavior in multi-turn conversational scenarios important for safety, including stereotyping-related harms. However, existing multi-turn social bias evaluations often rely on pre-specified or template-based user inputs that do not adapt to model responses and typically assume a fixed dialogue length in advance. In this paper, we study social bias dynamics in response-conditioned multi-turn interactions using a controlled evaluation protocol that generates follow-up user queries from the evolving dialogue history and allows evaluation over variable numbers of turns. Experimental results show that LLMs exhibit social bias even in coherent, response-conditioned multi-turn interactions, revealing late-emerging bias, non-monotonic bias patterns, and bias re-emergence. These results motivate evaluations that extend beyond fixed-turn, pre-scripted protocols. Our findings highlight the importance of analyzing social bias as a turn-level dynamic phenomenon.", "url": "https://wpnews.pro/news/dymt-esb-dynamic-multi-turn-evaluation-of-social-bias-in-user-llm-interactions", "canonical_source": "https://www.machinebrief.com/news/dymt-esb-dynamic-multi-turn-evaluation-of-social-bias-in-use-4p73", "published_at": "2026-09-17 04:00:00+00:00", "updated_at": "2026-09-17 06:54:55.024250+00:00", "lang": "en", "topics": ["ai-safety", "ai-ethics", "large-language-models", "ai-research"], "entities": ["DyMT-ESB", "arXiv"], "alternates": {"html": "https://wpnews.pro/news/dymt-esb-dynamic-multi-turn-evaluation-of-social-bias-in-user-llm-interactions", "markdown": "https://wpnews.pro/news/dymt-esb-dynamic-multi-turn-evaluation-of-social-bias-in-user-llm-interactions.md", "text": "https://wpnews.pro/news/dymt-esb-dynamic-multi-turn-evaluation-of-social-bias-in-user-llm-interactions.txt", "jsonld": "https://wpnews.pro/news/dymt-esb-dynamic-multi-turn-evaluation-of-social-bias-in-user-llm-interactions.jsonld"}}