Evolving Agent Teams：多智能体框架突破编码智能体局部最优

View PDF HTML (experimental)

Abstract:The rapid progress of software engineering agents drives interest in applying them to open-ended problems such as GPU kernel optimization. However, we find that existing coding agents converge to local optima on real-world, expert-written kernels (e.g., FlashInfer). We introduce Evolving Agent Teams, a multi-agent framework that escapes such local optima. It features (1) a central supervisor that diagnoses each local optimum and restructures a team of agents; and (2) a persistent file system that comprises a learning document, an evolution history, and per-team artifacts, grounding restructuring with accumulated experience. Under a matched token budget, Evolving Agent Teams consistently outperforms strong coding-agent baselines. With an extended token budget, Evolving Agent Teams generates a CUDA kernel that reaches 1.2x mean speedup over the production FlashInfer MLA paged decode kernel from scratch and wins on 36 of 47 production workloads from FlashInfer-Bench (0.86-0.99x on the remaining 11). It also demonstrates superior performance on tasks such as KDA prefill on AMD MI350X (3.01x over vLLM) and NVIDIA B200 (3.44x over FlashKDA), MoE expert-offload serving on Apple silicon (30.3 tokens/s decode for Qwen3.8-Flash-Next in MLX 4-bit on a 48 GB M4 Pro), and RoboDojo robot manipulation (62.1 on held-out seeds vs. 55.3 for the best leaderboard entry). Code is available at this https URL.

Subjects:	Artificial Intelligence (cs.AI)
Cite as:	arXiv:2610.36848 [cs.AI]
	(or arXiv:2610.36848v1 [cs.AI] for this version)
	https://doi.org/10.48550/arXiv.2610.36848 arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Shiyi Cao [view email]
[v1] Tue, 27 Oct 2026 08:46:21 UTC (199 KB)