Metraipolis › archive › dc_5cd39da22121
Investigating Assistant Bias in LLM User Simulators Using a Role Vector — arXiv:2609.00608
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.00608
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-17T06:41:17.559Z
- SHA-256 of the archived text
0f148e0d57d475de66d2eade9384eb048550312ff88b57c309c22f0096d0165e- Stable URL of this page
https://metraipolis.com/documento/dc_5cd39da22121
How to cite
Investigating Assistant Bias in LLM User Simulators Using a Role Vector — arXiv:2609.00608 — arXiv (en), https://arxiv.org/abs/2609.00608 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_5cd39da22121 (Metraipolis archive dc_5cd39da22121, sha256 0f148e0d57d475de66d2eade9384eb048550312ff88b57c309c22f0096d0165e)
BibTeX
@misc{dc5cd39da22121,
title = {Investigating Assistant Bias in LLM User Simulators Using a Role Vector — arXiv:2609.00608},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.00608},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_5cd39da22121; sha256 0f148e0d57d475de66d2eade9384eb048550312ff88b57c309c22f0096d0165e},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.00608. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
LLM-based user simulators are increasingly used to evaluate autonomous agents at scale, in place of costly human evaluations. Despite this promise, these simulators exhibit "assistant bias," a tendency to cooperate and pursue task goals. They rarely reproduce the frustration or disengagement that real users exhibit, compromising evaluation validity. Prior work outlines that this bias is baked in during model training, which role-playing prompts fail to override. We analyze this bias from model activations, extracting a user role vector by contrasting how the model represents user versus assistant perspectives on the same dialogue. We observe two findings: (i) the user direction is identifiable in activations, elicits user-like behaviors, and captures characteristics distinct from assistant traits; and (ii) although user-role activation associates with simulation realism and steering strengthens it, it can exaggerate user behaviors and override individual user profiles. Together, our findings provide a representation-level analysis of LLM user simulators, confirming that assistant bias is structurally identifiable and that user behavior can be directionally analyzed. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-02 in cs.CL). Categorie arXiv: Computation and Language (cs.CL) ; Human-Computer Interaction (cs.HC). Pagina del lavoro: https://arxiv.org/abs/2609.00608 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-17T06:33:38.817Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-09-02?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).