主动溯源门控：抑制多智能体辩论合成中的伪造共识
原标题：Towards Mitigating Fabricated Consensus: The Active Provenance Gate for Multi-Agent Debate Synthesis
AI 摘要
该论文提出「主动溯源门控」（Active Provenance Gate, APG），作为多智能体辩论（MAD）合成阶段之后的验证层，将辩论日志视为封闭证据世界，对每条主张进行审计并在违反约束时阻止发布，从而抑制「伪造共识」。在危机模拟中，自愈机制使困难场景下的平均数据溯源保真度提升一倍以上；人类研究中超过75%的用户在关键场景下更偏好明确报告失败的结果，尽管多数人认为基线系统的伪造共识更流畅。论文已被IEEE ICTAI 2026接收，由波兰国家科学中心资助。
正文节选
Towards Mitigating Fabricated Consensus: The Active Provenance Gate for Multi-Agent Debate SynthesisThanks: This work has been accepted for publication at the 38th IEEE International Conference on Tools with Artificial Intelligence (ICTAI 2026). © 2026 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new col