概率单纯形上的可组合解码：理论与实现

中文摘要 AI 辅助

大型语言模型的解码通常被视为一系列孤立的采样策略，对这些策略所引发的行为及其底层目标之间的关系的理论理解有限。我们将解码形式化为概率单纯形上关于下一词元分布的优化问题，在支持约束下平衡期望模型得分与正则化。这一视角通过选择正则化器和支持约束恢复了熟悉的解码方法；更重要的是，它使得能够通过在一个优化问题中组合分布偏好来构建新的解码器，而无需外部奖励、学习到的评判器或模型参数更新。我们引入了CompoSimplex，一个具有可配置支持规则、正则化原语和单纯形求解器的库，用于构建和评估组合解码器。我们在多个模型和推理任务上评估了标准采样器、单独的正则化器以及组合。我们的结果表明，组合可以实现单样本质量、多样本质量和多样性之间的权衡，而这些是单独的解码目标无法达到的。

英文摘要

Decoding for large language models is typically treated as a collection of isolated sampling strategies, with limited theoretical understanding of the behaviours they induce and how their underlying objectives relate. We formulate decoding as an optimisation problem over next-token distributions on the probability simplex, balancing expected model score against regularisation under support constraints. This view recovers familiar decoding methods through choices of regularisers and support constraints; more importantly, it enables new decoders to be constructed by composing distributional preferences within a single optimisation problem without external rewards, learned critics, or model parameter updates. We introduce CompoSimplex, a library with configurable support rules, regularisation primitives, and simplex solvers for constructing and evaluating compositional decoders. We evaluate standard samplers, individual regularisers, and compositions across multiple models and reasoning tasks. Our results show that compositions can realise trade-offs between single-sample quality, multi-sample quality, and diversity that are not attained by individual decoding objectives.

发表机构

• UCL Centre for AI（伦敦大学学院人工智能中心）

机构由 AI 辅助整理，请以论文原文为准。