arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2602.16793cs.LG
收藏

逃离认知陷阱:使用现成模型高效解决竞赛数学问题

Escaping the Cognitive Well: Efficient Competition Math with Off-the-Shelf Models

Princeton University(普林斯顿大学);Princeton Language and Intelligence(普林斯顿语言与智能)

Xingyu Dang, Rohit Agarwal, Rodrigo Porto, Anirudh Goyal, Liam H Fowl, Sanjeev Arora

AI总结:

提出一种推理流水线,利用现成模型以极低成本在IMO风格数学问题上达到最佳性能,通过猜想提取和上下文分离解决求解器-评分器流水线中的认知陷阱问题。

摘要:

在过去一年中,定制和未公开的数学推理模型在国际数学奥林匹克竞赛(IMO)中达到了金牌水平。随后,使用公开可用的模型通过大规模推理也报告了类似的性能,但成本高昂(例如,每个问题3000美元)。在这项工作中,我们提出了一种推理流水线,在IMO风格的数学问题上以平均推理成本比竞争方法低几个数量级的情况下实现了最佳性能,同时仅使用通用现成模型。我们的方法基于对求解器-评分器流水线中评分器失败的见解,我们称之为认知陷阱(迭代优化收敛到错误解,而求解器和流水线的内部评分器认为该解基本正确)。我们的流水线通过猜想提取来解决这些失败模式,其中候选引理从生成的解中分离出来,并在新环境(上下文分离)中与其否定形式一起独立验证。在IMO-ProofBench Advanced(PB-Adv)上,我们的流水线使用Gemini 3.0 Pro达到了67.1%的性能,每个问题的平均成本约为31美元。在评估时,这代表了PB-Adv上公开和未公开模型中的最先进水平,并且成功率是下一个最佳公开可访问流水线的两倍以上,而成本仅为其一小部分。

英文摘要:

In the past year, custom and unreleased math reasoning models reached gold medal performance on the International Mathematical Olympiad (IMO). Similar performance was then reported using large-scale inference on publicly available models but at prohibitive costs (e.g., 3000 USD per problem). In this work, we present an inference pipeline that attains best-in-class performance on IMO-style math problems at an average inference cost orders of magnitude below competing methods while using only general-purpose off-the-shelf models. Our method relies on insights about grader failure in solver-grader pipelines, which we call the Cognitive Well (iterative refinement converging to a wrong solution that the solver as well as the pipeline's internal grader consider to be basically correct). Our pipeline addresses these failure modes through conjecture extraction, wherein candidate lemmas are isolated from generated solutions and independently verified alongside their negations in a fresh environment (context detachment). On IMO-ProofBench Advanced (PB-Adv), our pipeline achieves 67.1 percent performance using Gemini 3.0 Pro with an average cost per question of approximately 31 USD. At the time of evaluation, this represented the state-of-the-art on PB-Adv among both public and unreleased models, and more than doubles the success rate of the next best publicly accessible pipeline, all at a fraction of the cost.

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。

↑