Adaptive Orchestration: Scalable Self-Evolving Multi-Agent Systems

TL;DR

This paper presents a scalable, self-evolving multi-agent system architecture that dynamically restructures itself with specialized sub-agents to improve task success and efficiency in large language model applications.

Contribution

It introduces a novel Self-Evolving Concierge System with a Dynamic Mixture of Experts approach, including a Meta-Cognition Engine and pruning mechanisms, to enhance scalability and stability.

Findings

Maintains high task success rates

Reduces token consumption compared to static swarms

Effectively manages capability gaps and resource constraints

Abstract

As Large Language Models (LLMs) are increasingly deployed as autonomous agents, they face a critical scalability bottleneck known as the "Generalization-Specialization Dilemma." Monolithic agents equipped with extensive toolkits suffer from context pollution and attention decay, leading to hallucinations. Conversely, static multi-agent swarms introduce significant latency and resource overhead. This paper introduces a Self-Evolving Concierge System, a novel architecture utilizing a Dynamic Mixture of Experts (DMoE) approach. Unlike recent self-improving agents that rewrite their own codebase, our system preserves stability by dynamically restructuring its runtime environment: "hiring" specialized sub-agents based on real-time conversation analysis. We introduce an asynchronous "Meta-Cognition Engine" that detects capability gaps, a Least Recently Used (LRU) eviction policy for resource constraints, and a novel "Surgical History Pruning" mechanism to mitigate refusal bias. Experimental results demonstrate that this architecture maintains high task success rates while minimizing token consumption compared to static agent swarms.