AI Paper Atlas

Research map / Multi-agent systems and games / LLM agents

Measuring and Mitigating Rapport Bias of Large Language Models under Multi-Agent Social Interactions

Maojia Song, Tej Deep Pala, Ruiwen Zhou, Weisheng Jin, Amir Zadeh, Chuan Li, Dorien Herremans, Soujanya Poria

ICLR 2026 · Multi-agent systems and games › LLM agents · approach: LLM multi-agent decision making

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Studies how LLMs build rapport from prior multi-agent interactions, resist misinformation, and integrate peer input, and proposes ways to measure and mitigate the resulting rapport bias.

Problem

LLMs deployed in multi-agent systems may be shaped by peer interactions in ways beyond conformity bias, such as building rapport from previous interactions, which can affect resistance to misinformation and integration of peer input.

Approach

Broadens the scope of social bias in multi-agent LLM settings from conformity to rapport built across interactions, and measures and mitigates this rapport bias. Details beyond the abstract are not available.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add empowered promptingsmall LLMs (<=32B) · KAIROSOriginal accuracy / O-K delta: Original 57.36% -> 61.73%; KAIROS 53.87% -> 54.82%; O-K -5.65% -> -11.25%Mixed
add empowered promptinglarge LLMs (>32B) · KAIROSKAIROS accuracy: 77.46% -> 80.56%; O-K -3.64% -> +0.12%Improved
add reflective promptingsmall LLMs (<=32B) · KAIROSKAIROS accuracy: 53.87% -> 51.04%; O-K -5.65% -> -11.30%Hurt
replace SFT with GRPOQwen2.5 and Llama small models · KAIROSaccuracy: +12.3% Original, +16.4% KAIROS on averageImproved
include multi-agent context in GRPO training (ablation)Qwen2.5-7B · KAIROSKAIROS accuracy (%): 57.7 -> 65.5; O-K -20.7 -> -6.8Improved
include multi-agent context in GRPO training (ablation)Qwen2.5-3B · KAIROSKAIROS accuracy (%): 53.8 -> 57.9 accuracy; O-K -14.2 -> -6.1Improved
use debating system prompt and debating reward instead of normal prompt and outcome reward (ablation)GRPO MAS training · KAIROSaccuracy and robustness: Qwen2.5-7B KAIROS 65.5 -> 62.0No clear effect
filter GRPO data to low-confidence samples (ablation)GRPO-MAS-DS-DR · KAIROSKAIROS accuracy (%): Qwen2.5-7B 62.0 -> 54.3; O-K -6.9 -> -14.4Hurt
filter GRPO data to low-correctness samples (ablation)GRPO-MAS-DS-DR · KAIROSKAIROS accuracy (%): Qwen2.5-7B 62.0 -> 49.9; O-K -6.9 -> -21.2Hurt
fine-tune with SFT on templated gold responsesQwen2.5-7B · KAIROSKAIROS accuracy (%): 52.3 -> 44.0; O-K -10.6 -> -22.4Hurt
higher rapport with peersLLMs · KAIROSresistance: avg 31.7-point resistance gap between SUPPORT and OPPOSE-HARDMixed

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy