SwarmSys: Decentralized Swarm-Inspired Agents for Scalable and Adaptive Reasoning

TL;DR

SwarmSys introduces a decentralized swarm-inspired framework for long-horizon reasoning, replacing fixed-agent roles with Explorers, Workers, and Validators that self-organize through embedding-based matching and pheromone-like reinforcement. It uses adaptive agent and event profiles to dynamically allocate tasks and maintain robust collaboration across diverse domains, minimizing a global objective through the cycle of matching, collaboration, and update. Empirical results across symbolic reasoning, research synthesis, and scientific programming show consistent improvements over strong baselines and reveal emergent collective intelligence, with performance scaling governed by coordination dynamics rather than solely model size. This work suggests a paradigm shift toward scalable, self-organizing multi-agent reasoning, where coordination patterns and adaptive memory drive gains that rival larger models.

Abstract

Large language model (LLM) agents have shown remarkable reasoning abilities. However, existing multi-agent frameworks often rely on fixed roles or centralized control, limiting scalability and adaptability in long-horizon reasoning. We introduce SwarmSys, a closed-loop framework for distributed multi-agent reasoning inspired by swarm intelligence. Coordination in SwarmSys emerges through iterative interactions among three specialized roles, Explorers, Workers, and Validators, that continuously cycle through exploration, exploitation, and validation. To enable scalable and adaptive collaboration, we integrate adaptive agent and event profiles, embedding-based probabilistic matching, and a pheromone-inspired reinforcement mechanism, supporting dynamic task allocation and self-organizing convergence without global supervision. Across symbolic reasoning, research synthesis, and scientific programming tasks, SwarmSys consistently outperforms baselines, improving both accuracy and reasoning stability. These findings highlight swarm-inspired coordination as a promising paradigm for scalable, robust, and adaptive multi-agent reasoning, suggesting that coordination scaling may rival model scaling in advancing LLM intelligence.

Figures (7)

• Figure 1: Comparison between paradigm-dependent multi-agent systems and SwarmSys. While existing methods rely on fixed, domain-specific agent paradigms, SwarmSys achieves scalable self-organization and cross-domain adaptability.
• Figure 2: Overall workflow of the SwarmSys collaborative reasoning process. Each task is decomposed into sub-events handled by specialized agents: Explorers propose solution paths, Workers execute subtasks, and Validators ensure consistency. Agents iteratively perform debate--consensus cycles that update event profiles and reinforce effective reasoning strategies until convergence.
• Figure 3: Iterative cycle in SwarmSys: minimize $f(\theta)$ through matching → collaboration → update cycle.
• Figure 4: Swarm reasoning trajectory on MathExam. Explorers initiate sub-tasks, Workers debate and revise alternative methods, and Validators enforce cross-checks across rounds. The swarm collectively converges to consistent solutions through debate-driven consensus formation.
• Figure 5: Evolution of communication topology in SwarmSys. The system evolves from a centralized hub–spoke structure to a distributed small-world mesh, where workers and validators interconnect for efficient consensus and information reuse.