优化大模型推理中验证资源分配，减少冗余调用。
Adaptive Test-Time Compute Allocation via Learned Heuristics over Categorical Structure
- 基于状态结构设计选择性验证框架，动态筛选待验证中间结果。
- 在MATH基准上准确率超越最佳N选一，验证调用减少44%。
- 适合追求高效推理的开发者，尤其关注验证成本控制场景。
测试时计算已成为大语言模型推理进步的主要驱动力，但日益受制于昂贵的验证开销。在许多推理系统中，大量验证调用消耗在冗余或无前景的中间假设上。本文研究验证成本受限下的推理问题，探讨如何在中间状态间合理分配验证资源。提出一种基于状态级别的选择性验证框架，结合（i）结构化动作接口上的确定性可行性门控，（ii）融合学习状态距离与残差评分的预验证排序，以及（iii）基于局部不确定性的验证调用自适应分配。相较于解级最优N选一或均匀中间验证，本方法将验证集中在信息量最高的状态。在MATH基准上，该方法在准确率上优于best-of-N、多数投票和束搜索，同时验证调用减少44%。
原文摘要 · Abstract (English)
Test-time computation has become a primary driver of progress in large language model (LLM) reasoning, but it is increasingly bottlenecked by expensive verification. In many reasoning systems, a large fraction of verifier calls are spent on redundant or unpromising intermediate hypotheses. We study reasoning under a \emph{verification-cost-limited} setting and ask how verification effort should be allocated across intermediate states. We propose a state-level selective verification framework that combines (i) deterministic feasibility gating over a structured move interface, (ii) pre-verification ranking using a hybrid of learned state-distance and residual scoring, and (iii) adaptive allocation of verifier calls based on local uncertainty. Unlike solution-level best-of-$N$ or uniform intermediate verification, our method distributes verification where it is most informative. On the \textsc{MATH} benchmark, our approach achieves higher accuracy than best-of-$N$, majority voting, and beam search while using 44\% fewer verifier calls.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。