交互税：当多智能体团队中的沟通消除多样性时
The Interaction Tax: When Communication Erases Diversity in Multi-Agent Teams
浏览论文内容
中文总结 AI 辅助
该研究提出“交互税”概念，发现多智能体LLM的完整方案交互会消除多样性，独立提议生成可避免收敛崩溃，交互仅在共享恰当信息时才有益。
中文摘要 AI 辅助
多智能体大语言模型（LLM）的交互究竟是有益还是有害？部分研究表明辩论（Du等人，2024）、批评循环（Chen等人，2025）以及多智能体混合合成（Wang等人，2025）可带来性能提升；而另一些研究则发现，在预算相等的情况下，交互会增加成本却无法提升质量（Tran & Kiela，2026；Xu等人，2026；Jarrett等人，2025），或是独立采样已能捕捉多智能体的增益（Li等人，2024）。我们认为这种矛盾部分源于一个缺失的区分：并非所有多智能体沟通的性质都相同。不同模型族会找到结构不同的解决方案，但当智能体读取彼此的完整输出时，其提议在一轮内就会收敛，从而消除了使用多个模型的核心动因——多样性，我们将此称为“交互税”。我们在匹配预算下测试了11个验证器评分的优化任务，发现完整方案交互是一种弱默认设置：独立提议生成可避免这种收敛崩溃；完整方案交互主要使智能体倾向于靠近其看到的第一个解决方案，而非尝试不同方法；而批评仅在被违反的规则易于被LLM发现和修正时才会有帮助。这些结果表明，多智能体性能更多取决于智能体交换的信息，而非智能体数量，且仅当智能体在恰当时间共享恰当信息时，交互才会发挥作用。
英文摘要
Does multi-agent LLM interaction help or hurt? Some work reports gains from debate (Du et al., 2024), critique loops (Chen et al., 2025), and mixture-of-agents synthesis (Wang et al., 2025), while other work finds that interaction adds cost without improving quality under equal budgets (Tran & Kiela, 2026; Xu et al., 2026; Jarrett et al., 2025), or that independent sampling already captures multi-agent gains (Li et al., 2024). We argue this contradiction partly reflects a missing distinction, because not all multi-agent communication is equal. Different model families find structurally different solutions, but when agents read each other's complete outputs, their proposals converge within one round, erasing the diversity that motivates using multiple models. We call this the interaction tax. We test 11 verifier-scored optimization tasks under matched budgets and find that full-solution interaction is a weak default. Independent proposal generation avoids this collapse. Full-solution interaction mainly makes agents stay close to the first solution they see instead of trying different approaches, and critique helps only if the violated rule is easy for the LLM to find and fix. These results suggest that multi-agent performance depends less on the number of agents than on the information they exchange, and interaction helps only when agents share the right information at the right time.
发表机构
- University of Chicago（芝加哥大学）
机构由 AI 辅助整理，请以论文原文为准。
