Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Beyond Consensus: Trace-Level Synthesis in Mixture of Agents

About

When multiple LLM agents solve the same problem, standard practice compresses each agent's reasoning into a majority vote or layered synthesis, treating agreement as the finish line. We show this is unnecessarily lossy: an LLM aggregator that reads complete reasoning traces recovers correct solutions even when agents unanimously agree, with beneficial corrections consistently outweighing harmful ones -- the \emph{aggregation paradox}. Majority voting has a ceiling that perturbation diversity does not raise (error correlations are identical); the aggregator's gain comes from trace-level complementarity, assembling correct intermediate steps from minority chains that voting discards. These findings motivate Self-Consistent Mixture of Agents which generates trace diversity through semantic-preserving input perturbations, safeguards the majority via anchored refinement with provable non-degradation guarantees, and always synthesizes -- never gates on consensus. A single model with perturbation-induced trace variation outperforms heterogeneous model pools across structured reasoning, PhD-level science, competition mathematics, and competitive programming. The unit of aggregation should be the reasoning trace, not the answer.

Shreyas Fadnavis, Praitayini Kanakaraj, Felix Wyss• 2026

Related benchmarks

TaskDatasetResultRank
Question AnsweringMMLU--
13
Code GenerationLCB-Hard (171)
Accuracy62.6
8
Question AnsweringBBH-3 296
Accuracy86.5
8
Question AnsweringGPQA (198)
Accuracy73.2
8
Question AnsweringAIME 90
Accuracy91.1
8
Showing 5 of 5 rows

Other info

Follow for update