Reaching Agreement Among Reasoning LLM Agents

2025-12-23

AI Summary

Current multi-LLM agent coordination lacks a formal consensus foundation, relying instead on static heuristics (e.g., fixed-round voting or barrier synchronization), leading to resource inefficiency, high latency, and ad hoc protocol designs. This work pioneers the application of distributed consensus theory to collaborative inference among stochastic reasoning agents. We propose the first distributed consensus framework tailored for such agents: (i) a formally specified multi-agent refinement model grounded in operational semantics, and (ii) Aegean—a novel consensus protocol supporting dynamic, incremental quorum detection and early termination. We implement Aegean-Serve, a consensus-aware serving engine. Evaluated on four mathematical reasoning benchmarks, Aegean reduces inference latency by 1.2×–20× while preserving answer quality at ≥97.5%. Crucially, the framework provides provable safety and liveness guarantees—addressing correctness and progress under asynchrony and agent unreliability.

Technology Category

• Multiagent Systems: Agreement, Argumentation & Negotiation
• Cognitive Modeling & Cognitive Systems: Conceptual Inference and Reasoning
• Reasoning under Uncertainty: Sequential Decision Making

Application Category

• Search and Retrieval-Augmented AI: Agentic search
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Multi-agent systems have extended the capability of agentic AI. Instead of single inference passes, multiple agents perform collective reasoning to derive high quality answers. However, existing multi-agent orchestration relies on static heuristic workflows such as fixed loop limits and barrier synchronization. These ad-hoc approaches waste computational resources, incur high latency due to stragglers, and risk finalizing transient agreements. We argue that reliable multi-agent reasoning requires a formal foundation analogous to classical distributed consensus problem. To that end, we propose a formal model of the multi-agent refinement problem. The model includes definitions of the correctness guarantees and formal semantics of agent reasoning. We then introduce Aegean, a consensus protocol designed for stochastic reasoning agents that solves multi-agent refinement. We implement the protocol in Aegean-Serve, a consensus-aware serving engine that performs incremental quorum detection across concurrent agent executions, enabling early termination when sufficient agents converge. Evaluation using four mathematical reasoning benchmarks shows that Aegean provides provable safety and liveness guarantees while reducing latency by 1.2--20$ imes$ compared to state-of-the-art baselines, maintaining answer quality within 2.5%. Consistent gains across both local GPU deployments and commercial API providers validate that consensus-based orchestration eliminates straggler delays without sacrificing correctness.

Problem

• Formalizing multi-agent reasoning as a distributed consensus problem
• Designing a consensus protocol for stochastic reasoning agents
• Reducing latency and ensuring correctness in multi-agent systems

Innovation

• Formal consensus protocol for stochastic reasoning agents
• Incremental quorum detection enabling early termination
• Provable safety and liveness guarantees reducing latency significantly