COMED: The Missing Middle Between Routing and Collaboration in Multi-LLM Inference

COMED：多大模型推理中路由与协作之间的“缺失环节”

Abstract: No single Large Language Model (LLM) is uniformly reliable across queries, motivating multi-model inference systems that either route among models or combine their outputs. However, routing stops after selecting an initial model, while dense collaboration invokes peers on every query. We show that collaboration is non-monotonic: peers can recover failures that no model solves alone, but can also corrupt initially correct answers.

摘要： 没有单一的大语言模型（LLM）能够在所有查询中保持一致的可靠性，这促使了多模型推理系统的出现，这些系统要么在模型间进行路由，要么结合它们的输出。然而，路由在选择初始模型后即停止，而密集协作则会在每次查询时都调用同伴模型。我们研究发现，协作是非单调的：同伴模型可以修复单一模型无法解决的错误，但也可能破坏最初正确的答案。

We introduce COMED (Controlled Model Escalation for Multi-LLM Deliberation), a post-anchor controller for selective cross-model collaboration. COMED uses anchor self-consistency, router margin, and a lightweight peer probe to accept confident answers, verify ambiguous cases, and escalate only when collaboration is likely beneficial. We formalize this trade-off with a rescue-harm decomposition showing that selective collaboration improves when rescued errors outweigh collaboration-induced harms.

我们引入了 COMED（多大模型审议的受控模型升级），这是一种用于选择性跨模型协作的锚点后控制器。COMED 利用锚点自洽性、路由余量和轻量级同伴探测来接受高置信度答案、验证模糊案例，并仅在协作可能带来收益时才进行升级。我们通过“救援-伤害”分解法将这种权衡形式化，表明当挽救的错误超过协作带来的伤害时，选择性协作的效果会得到提升。

Across medical, scientific, and general reasoning benchmarks, COMED improves fixed and routed anchors in all 16 open-weight settings, with gains up to +10.7 percentage points on MedQA while invoking fewer models and using fewer decoded tokens than dense collaboration. On HLE with frontier models, COMED improves GPT-5.5 from 23.1% to 28.1%, outperforming dense collaboration and achieving the best results.

在医学、科学和通用推理基准测试中，COMED 在所有 16 种开源权重设置下均改进了固定和路由锚点，在 MedQA 上提升了高达 10.7 个百分点，同时相比密集协作调用了更少的模型并使用了更少的解码 Token。在基于前沿模型的 HLE 测试中，COMED 将 GPT-5.5 的表现从 23.1% 提升至 28.1%，超越了密集协作并取得了最佳结果。