Qian Zhang, Zhongxiang Dai, Yao Shu, Mingze Kong, Zhiwei Shang, Zhi Hong, Jiahang Sun, Xiangyi Wang
We have not lifted any functions out of this paper's repositories yet, so there is nothing we have run. If it links a repository, it is listed below.
Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.
Large Language Models (LLMs) have achieved substantial success in real-world applications, particularly as the cognitive backbone of Multi-Agent Systems (MAS) for orchestrating complex workflows. Since many deployments preclude workflow modifications while MAS performance is highly prompt-sensitive, prompt optimization becomes a critical strategy for improvement. However, real-world prompt optimization for MAS is impeded by three key challenges: (1) the need of sample efficiency due to prohibitive evaluation costs, (2) topology-induced coupling among prompts, and (3) the combinatorial explosion of the search space. To address these challenges, we introduce MASPOB (Multi-Agent System Prompt Optimization via Bandits), a novel sample-efficient framework based on bandits. By leveraging Upper Confidence Bound (UCB) to quantify uncertainty, the bandit framework balances exploration and exploitation, maximizing gains within a strictly limited budget. To handle topology-induced coupling, MASPOB integrates Graph Neural Networks (GNNs) to capture structural priors, learning topology-aware representations of prompt semantics. Furthermore, it employs coordinate ascent to decompose the optimization into univariate sub-problems, reducing search complexity from exponential to linear. Extensive experiments across diverse benchmarks demonstrate that MASPOB achieves stateof-the-art performance, consistently outperforming existing baselines. Our code is available at https://github.com/HZ1008/MASPOB.
The same record, over MCP at https://syntology.ai/mcp:
get_harvested_code_for_paper("2603.02630")
get_code_for_paper("2603.02630")
have("2603.02630")
Connect an agent — have() is free.