AI Paper Atlas

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

Thinking as Society: Multi-Social-Agent Self-Distillation for Multimodal Misinformation Detection

Yifei Gao, Ning Xu, Wenhui Li, Hongshuo Tian, Lanjun Wang, Anan Liu

ICLR 2026 · Multi-agent systems and games › LLM agents · approach: Self-evolving language agent learning

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.

Proposes a multi-social-agent self-distillation framework for multimodal misinformation detection, using MLLM agents that simulate diverse social viewpoints.

Problem

Multimodal misinformation detection in mixed-source scenarios requires robust reasoning over social complexity and diverse forgery types; single-agent MLLM methods give only limited single-view analysis.

Approach

Uses multiple social agents to provide diverse analytic perspectives and distills their collective reasoning into a single model via self-distillation, addressing the single-view versus costly multi-agent trade-off (abstract is truncated, so details beyond this are not confirmed).

Measured effects reported in the paper

What was changedWhereMeasuredEffect
train Qwen2-VL on SCoT data with SCPOQwen2-VL-7B · MFC-Bench overallaccuracy: 57.24 -> 67.15Improved
add social correction value scaling to ORPO loss (SCPO vs ORPO) (ablation)SCoT-trained Qwen2-VL-7B · MFC-Bench overallaccuracy: 66.30 -> 67.15Improved
add preference loss on top of SFT (ORPO vs SFT) (ablation)SCoT-trained Qwen2-VL-7B · MFC-Bench overallaccuracy: 64.20 -> 66.30Improved
add DPO after SFT (ablation)SCoT-trained Qwen2-VL-7B · MFC-Bench overallaccuracy: 64.20 -> 57.98Hurt
SFT on SCoT data instead of self-consistency prompting (ablation)Qwen2-VL-7B · MFC-Bench overallaccuracy: 61.63 -> 64.20Improved
self-consistency prompting (ablation)Qwen2-VL-7B · MFC-Bench OOCaccuracy: 63.85 -> 52.60Hurt
self-consistency prompting (ablation)Qwen2-VL-7B · MFC-Bench overallaccuracy: 57.46 -> 61.63Improved
apply MMD-Agent multi-agent frameworkQwen2-VL-7B · MMFakeBench (closed prompting)accuracy: 41.2 -> 35.6Hurt
apply MMD-Agent multi-agent frameworkLLaVA-Next-34B · MMFakeBench (closed prompting)accuracy: 46.6 -> 40.5Hurt
apply MMD-Agent multi-agent frameworkQwen2-VL-7B · MMFakeBench (closed prompting)F1: 31.8 -> 32.8Improved
train with SCPOQwen2-VL-7B · MMFakeBench (open prompting, Top-1)accuracy: 41.0 -> 59.2Improved
train with SCPOQwen2-VL-7B · MFC-Bench reasoning qualitymisleadingness: 2.82 -> 2.51Improved

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