arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2605.06339cs.AI
收藏

控制器类别选择的制度理论:用于LLM行动决策

A Regime Theory of Controller Class Selection for LLM Action Decisions

University of Glasgow(格拉斯哥大学);University College London(伦敦大学学院);UESTC(电子科技大学)

Zhaoyang Jiang, Zhizhong Fu, Yunsoo Kim, Jiacong Mi, Zicheng Li, Xuanqi Peng, Honghan Wu

AI总结:

本文提出一种制度理论,通过分析有限样本下的实例不确定性信号,指导LLM在不同输入下选择合适的控制器类别,从而在有限样本中实现最优决策。

摘要:

部署的语言模型和视觉-语言模型必须在每个输入上决定是否直接回答、检索证据、委托给更强的模型或 abstain。与常见的单调性直觉相反,在有限样本中,更高的输入表达性不总是有益:在相同的严格交叉验证下,不同的基准偏好不同的控制器类别。这反映了实例级不确定性信号的有限样本限制,可以在分布依赖的尺度上耗尽。我们将控制器组织成一个包含四个类别的嵌套晶格:固定动作、分区路由器、实例级控制器和先验门控控制器,按复杂性排序。我们证明了一种制度理论,将三个可数据估计的瓶颈转化为类别选择:在最佳固定动作之上可以实现多少改进,实例级控制器是否有足够的样本做出可靠决策,以及当实例级信号不可靠时,粗粒度分区路由器能恢复多少改进。所得到的伯恩斯坦紧阈值具有匹配的信息论下界,严格的嵌套交叉验证可证明选择接近最佳的类别。在SMS-Spam、HallusionBench、A-OKVQA和FOLIO上,预测的类别与经验赢家一致;在TextVQA上,当OCR令牌提供无标签的预测时间先验时,先验门控控制器获胜。代码可在https://github.com/Anonymous-Awesome-Submissions/Regime-Theory上获得。

英文摘要:

Deployed language and vision-language models must decide, on each input, whether to answer directly, retrieve evidence, defer to a stronger model, or abstain. Contrary to the common monotonicity intuition, greater per-input expressivity is not uniformly beneficial in finite samples: under identical strict cross-validation, different benchmarks prefer different controller classes. This reflects a finite-sample limitation of instance-level uncertainty signals, which can be exhausted at a distribution-dependent scale. We organize controllers into a nested lattice of four classes: fixed actions, partition routers, instance-level controllers, and prior-gated controllers, ordered by complexity. We prove a regime theory that turns three data-estimable bottlenecks into a class choice: how much improvement is possible beyond the best fixed action, whether there are enough samples for instance-level controllers to make reliable decisions, and how much improvement a coarse partition router can recover when instance-level signal is unreliable. The resulting Bernstein-tight threshold has a matching information-theoretic lower bound, and strict nested cross-validation provably selects a near-best class. Across SMS-Spam, HallusionBench, A-OKVQA, and FOLIO, the predicted class matches the empirical winner; the prior-gated controller wins on TextVQA when OCR tokens supply a label-free prediction-time prior. Code is available at https://github.com/Anonymous-Awesome-Submissions/Regime-Theory.

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。

↑