arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2603.15968cs.AIcs.CLcs.LGcs.MA
收藏

MAC:多智能体宪法学习

MAC: Multi-Agent Constitution Learning

Mohamed bin Zayed University of Artificial Intelligence(莫扎德人工智能大学);Indraprastha Institute of Information Technology Delhi(德里印度教教派信息科技学院);Google DeepMind(谷歌DeepMind)

Rushil Thareja, Gautam Gupta, Francesco Pinto, Nils Lukas

AI总结:

本文提出MAC多智能体宪法学习方法,通过结构化提示优化提升LLM控制效果,实现可解释的规则集生成,优于现有提示优化方法,并在PII标注任务中表现优异。

摘要:

本文提出MAC多智能体宪法学习方法,通过结构化提示优化提升LLM控制效果,实现可解释的规则集生成,优于现有提示优化方法,并在PII标注任务中表现优异。

英文摘要:

Constitutional AI is a method to oversee and control LLMs based on a set of rules written in natural language. These rules are typically written by human experts, but could in principle be learned automatically given sufficient training data for the desired behavior. Existing LLM-based prompt optimizers attempt this but are ineffective at learning constitutions since (i) they require many labeled examples and (ii) lack structure in the optimized prompts, leading to diminishing improvements as prompt size grows. To address these limitations, we propose Multi-Agent Constitutional Learning (MAC), which optimizes over structured prompts represented as sets of rules using a network of agents with specialized tasks to accept, edit, or reject rule updates. We also present MAC+, which improves performance by training agents on successful trajectories to reinforce updates leading to higher reward. We evaluate MAC on tagging Personally Identifiable Information (PII), a classification task with limited labels where interpretability is critical, and demonstrate that it generalizes to other agentic tasks such as tool calling. MAC outperforms recent prompt optimization methods by over 50%, produces human-readable and auditable rule sets, and achieves performance comparable to supervised fine-tuning and GRPO without requiring parameter updates.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑