Files
nexus/wiki/concepts/Adversarial-Debate-Pattern.md

35 lines
1.2 KiB
Markdown
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
---
title: "Adversarial Debate Pattern"
type: concept
tags: []
sources:
- multi-agent-system-reliability
last_updated: 2026-04-28
---
# Adversarial Debate Pattern
## 定义
多智能体系统的对抗式辩论模式——一个Agent提出方案另一个Agent攻击反驳由第三个Agent裁判决定胜负。核心是用外部批评者和评判者模拟人类的"恐惧"动机。
## 角色
- **Generator**"Here is my plan."(生成方案)
- **Critic**"Here are 3 reasons why that plan sucks."(扮演魔鬼代言人)
- **Judge**"The Critic is right. Fix it."(裁判/主持人)
## 核心洞察
LLM是"Yes-Men",一旦开始写作很少自我纠正——需要一个指定的反对者来打破这种惯性。
## 关键机制
- 三方应使用**不同模型**(不同训练/微调/提示),多样性有益
- 顺序执行+循环特性导致速度可能非常慢
- Agent可能陷入无限辩论——可使用**Watchdog**(确定性代码)在时间/次数超阈值时打破循环
## 适用场景
- 安全分析Security Analysis
- 代码审查Code Review
- 高风险内容审核High-Stakes Content Moderation
## 来源
- [[multi-agent-system-reliability]]