今日速览
| 序号 | 标题 | 来源 | 日期 | 主题 | 推荐等级 |
|---|---|---|---|---|---|
| 1 | When Agents Disagree: Bayesian Backward Reasoning as a Label-Free Anchor for Multi-Agent Collective Decision-Making | arXiv | 2026-09-10 | AI-Agent | 高 |
| 2 | COBRA-Skills: Contextual Bandit-Guided Evolution for Agent Skill Optimization | arXiv | 2026-09-10 | RAG | 高 |
重点论文与技术动态
1. When Agents Disagree: Bayesian Backward Reasoning as a Label-Free Anchor for Multi-Agent Collective Decision-Making
- 来源：arXiv
- 日期：2026-09-10
- 作者/机构：Ken Chen, Wei Wang, Sachith Seneviratne, Hansani Weeratunge, Saman Halgamuge
- 主题标签：
AI-Agent,arXiv - 推荐等级：高
- 分类：cs.AI, cs.MA
一句话结论
本研究通过贝叶斯反向推理构建标签自由锚点，显著提升多代理集体决策性能，尤其在代理冲突场景下表现突出。
核心内容
- 现有集体决策方法（如投票、LLM裁判）依赖前向推理，易继承共享错误。
- 提出反向后验通过贝叶斯反向推理，提供与前向不同的因子化近似，缓解相关错误。
- 基于Jensen-Shannon散度的交叉路径一致性排名代理，实现MinJS、FwdJS和LogLin三种策略，LogLin在代理冲突子集表现最佳。
方法与数据
- 方法：贝叶斯反向推理与Jensen-Shannon散度计算交叉路径一致性。
- 数据：DDXPlus数据集，五个LLM骨干评估。
价值判断
- 值得关注：反向后验作为互补锚点，在代理冲突时性能提升显著，提供标签自由决策支持。
- 可复用点：交叉路径一致性机制可推广至其他多代理系统，无需额外标签。
- 局限/待核查：反向后验独立准确性较低，需标签数据进行轻量级两阶段校准以优化。
摘要
When multiple LLM agents yield conflicting answers, the decision-making process dictates whether agent diversity improves performance or merely compounds shared errors. Existing collective decision-making methods, including voting, electoral rules, and LLM judges, rely on forward reasoning: they map evidence to labels in one direction. Although these methods can combine diverse forward traces, they still aggregate estimates that share this evidence-to-label factorization and can inherit correlated errors within the forward pool. We therefore construct a reverse posterior for each instance through Bayesian backward reasoning from an explicit likelihood. The forward and reverse posteriors provide differently factorized approximations of the underlying posterior. Because estimates from different factorizations may tend to share the same error less often, we use Jensen-Shannon divergence to rank agents by cross-path consistency. This cross-path consistency signal underlies three strategies: hard selection (MinJS), soft reweighting (FwdJS), and log-linear fusion (LogLin). Evaluated on DDXPlus across five LLM backbones, our proposed strategies show consistent improvements: MinJS outperforms random selection across all backbones, FwdJS generally improves over the strongest baseline, and LogLin achieves the best performance among the evaluated methods, with its largest gains on the subset where the agents disagree. Despite its weaker standalone accuracy, the reverse posterior serves as a more useful anchor than forward-only alternatives, providing complementary information for collective decision-making. When labeled data are available, a lightweight two-stage calibration can further refine the reverse anchor and improve aggregation performance.2. COBRA-Skills: Contextual Bandit-Guided Evolution for Agent Skill Optimization
- 来源：arXiv
- 日期：2026-09-10
- 作者/机构：Pingchen Lu, Xiangyi Wang, Xiang Li, Jie Mao, Zikun Qu, Junfeng Luo
- 主题标签：
RAG,arXiv - 推荐等级：高
- 分类：cs.AI
一句话结论
COBRA-Skills通过上下文带状引导优先级分配与证据驱动技能进化，实现了高效技能优化，在六种异构代理基准上性能最优，优化成本比SkillOpt降低55-58%。
核心内容
- 将技能优化形式化为预算化的序列优化问题，构建动态演化的候选空间，减少对昂贵执行评估的依赖。
- 耦合上下文带状引导优先级分配与证据驱动技能进化，选择性分配评估资源到有前景候选者，并从执行反馈中持续优化技能群体。
- 在六种异构代理基准和三种目标模型上，性能显著优于对比方法，优化成本降低55-58%，每基准仅需50个唯一优化示例。
方法与数据
- 方法：上下文带状引导优先级分配与证据驱动技能进化。
- 数据：摘要未明确。
价值判断
- 值得关注：优化成本大幅降低55-58%，性能提升显著，且对代理框架变化保持鲁棒性。
- 可复用点：框架可推广至不同LLM代理任务，减少对大量任务数据的依赖，支持目标模型自身用于技能生成与优化。
- 局限/待核查：摘要未明确。