MetaGen: Self-Evolving Roles and Topologies for Multi-Agent LLM Reasoning
This addresses inefficiencies in multi-agent LLM systems for complex tasks, offering a more flexible and cost-effective approach, though it is incremental as it builds on existing multi-agent methods.
The paper tackles the problem of rigid role libraries and interaction topologies in multi-agent LLM systems, which cause task mismatches and high inference costs, by introducing MetaGen, a training-free framework that dynamically adapts roles and topologies at inference time, improving accuracy and cost tradeoffs on benchmarks like code generation and multi-step reasoning.
Large language models are increasingly deployed as multi-agent systems, where specialized roles communicate and collaborate through structured interactions to solve complex tasks that often exceed the capacity of a single agent. However, most existing systems still rely on a fixed role library and an execution-frozen interaction topology, a rigid design choice that frequently leads to task mismatch, prevents timely adaptation when new evidence emerges during reasoning, and further inflates inference cost. We introduce MetaGen, a training-free framework that adapts both the role space and the collaboration topology at inference time, without updating base model weights. MetaGen generates and rewrites query-conditioned role specifications to maintain a controllable dynamic role pool, then instantiates a constrained execution graph around a minimal backbone. During execution, it iteratively updates role prompts and adjusts structural decisions using lightweight feedback signals. Experiments on code generation and multi-step reasoning benchmarks show that MetaGen improves the accuracy and cost tradeoff over strong multi-agent baselines.