The Cost of Consensus: Isolated Self-Correction Prevails Over Unguided Homogeneous Multi-Agent Debate
Multi-agent debate, where teams of LLMs iteratively exchange rationales and vote on answers, is widely deployed under the assumption that peer review filters hallucinations. Yet the failure dynamics of homogeneous debate remain poorly understood, therefore we report findings from a controlled empirical study of teams of $N{=}10$ homogeneous agents (Qwen2.5-7B, Llama-3.1-8B, Ministral-3-8B) across $R{=}3$ debate rounds on two high-difficulty benchmarks (GSM-Hard and MMLU-Hard). We compare peer debate against isolated self-correction and a stochastic noise control that injects rationales from unrelated problems. We decompose debate failure into three model-dependent pathways: sycophantic conformity, where agents uncritically adopt majority answers (modal adoption up to 85.5%); contextual fragility, where peer rationales destabilize previously correct reasoning (vulnerability rate up to 70.0%); and consensus collapse, where plurality voting discards correct answers already present in the generation pool (oracle gap up to 32.3 percentage points). Ablations over communication density ($K \in \{2,4,9\}$) and sampling temperature ($T \in \{0.4, 0.7\}$) show that conformity reaches high levels at minimal peer exposure ($K{=}2$) and intensifies with greater initial diversity. Across all configurations, debate consumes 2.1-3.4$\times$ more tokens (up to 28,631 tokens per problem) than self-correction for equal or lower accuracy. Our results indicate that, within the 7-8B parameter class, homogeneous teams without structured roles do not benefit from unguided peer exchange, and that isolated self-correction consistently offers a more favorable cost-accuracy tradeoff.
Code (0)
등록된 구현이 없습니다.
Similar Papers 제목 키워드 기반
Divide-and-Conquer Meets Consensus: Unleashing the Power of Functions in Code Generation
Despite recent progress made by large language models in code generation, they still struggle with programs that meet complex requirements. Recent work utilizes plan-and-solve decomposition to decrease the complexity and…
Code GenerationHumanEvalMathmbppWhen Can LLMs Actually Correct Their Own Mistakes? A Critical Survey of Self-Correction of LLMs
Self-correction is an approach to improving responses from large language models (LLMs) by refining the responses using LLMs during inference. Prior work has proposed various self-correction frameworks using different so…
SurveyStable Answers, Unfinished Reasoning: Why Self-Consensus Is Not a Safe Early-Exit Signal
A natural way to cut reasoning-model inference cost is to repeatedly probe a single partial trajectory for its current answer and stop once probes agree -- self-consensus. We ask whether any such rule is both safe and to…
CEC-Zero: Zero-Supervision Character Error Correction with Self-Generated Rewards
Large-scale Chinese spelling correction (CSC) remains critical for real-world text processing, yet existing LLMs and supervised methods lack robustness to novel errors and rely on costly annotations. We introduce CEC-Zer…
Reinforcement LearningSemantic SimilarityMoral Machine or Tyranny of the Majority?
With Artificial Intelligence systems increasingly applied in consequential domains, researchers have begun to ask how these systems ought to act in ethically charged situations where even humans lack consensus. In the Mo…
Autonomous VehiclesFairness