EMS: Multi-Agent Voting via Efficient Majority-then-Stopping

arXiv cs.AI / 4/6/2026

💬 OpinionSignals & Early TrendsIdeas & Deep AnalysisModels & Research

共有:

Key Points

The paper argues that conventional majority voting in multi-agent systems wastes computation because many agents continue reasoning even after a majority consensus is already reachable.
It introduces Efficient Majority-then-Stopping (EMS), which casts voting as a reliability-aware agent scheduling problem and stops as soon as a majority decision can be formed.
EMS relies on Agent Confidence Modeling (ACM) to estimate per-agent reliability from historical performance and semantic similarity.
It uses Adaptive Incremental Voting (AIV) to select agents sequentially and Individual Confidence Updating (ICU) to revise reliability estimates as more agent outputs are incorporated.
Experiments on six benchmarks show EMS reduces the average number of invoked agents by 32%, improving reasoning efficiency without changing the majority-voting aggregation goal.

Abstract

Majority voting is the standard for aggregating multi-agent responses into a final decision. However, traditional methods typically require all agents to complete their reasoning before aggregation begins, leading to significant computational overhead, as many responses become redundant once a majority consensus is achieved. In this work, we formulate the multi-agent voting as a reliability-aware agent scheduling problem, and propose an Efficient Majority-then-Stopping (EMS) to improve reasoning efficiency. EMS prioritizes agents based on task-aware reliability and terminates the reasoning pipeline the moment a majority is achieved from the following three critical components. Specifically, we introduce Agent Confidence Modeling (ACM) to estimate agent reliability using historical performance and semantic similarity, Adaptive Incremental Voting (AIV) to sequentially select agents with early stopping, and Individual Confidence Updating (ICU) to dynamically update the reliability of each contributing agent. Extensive evaluations across six benchmarks demonstrate that EMS consistently reduces the average number of invoked agents by 32%.