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 2602.15136stat.MLcs.LG
收藏

通用先验:通过贝叶斯推断和预训练解决经验贝叶斯问题

Universal priors: solving empirical Bayes via Bayesian inference and pretraining

Courant Institute of Mathematical Sciences, NYU(纽约大学Courant数学科学研究所);Center for Data Science, NYU(纽约大学数据科学中心);Department of EECS, MIT(麻省理工学院电子工程与计算机科学系)

Nick Cannella, Anzo Teh, Yanjun Han, Yury Polyanskiy

AI总结:

本文从理论上证明,在合成数据上预训练的Transformer能通过后验收缩自适应未知测试分布,实现经验贝叶斯问题的近最优遗憾界。

摘要:

我们从理论上证明了[Teh等人,2025]最近的实证发现:在合成生成数据上预训练的Transformer在经验贝叶斯(EB)问题上表现出色。我们采取间接方法:不分析模型架构或训练动态,而是探究为什么在预设训练分布下训练的预训练贝叶斯估计器能够适应任意测试分布。聚焦于泊松EB问题,我们识别出通用先验的存在,使得在这些先验下训练能够获得均匀地关于所有测试分布的近最优遗憾界$\widetilde{O}(\frac{1}{n})$。我们的分析利用了贝叶斯统计中后验收缩的经典现象,表明预训练Transformer正是通过后验收缩适应未知测试分布。这一视角还解释了长度泛化现象(测试序列长度超过训练长度),因为模型使用广义后验执行贝叶斯推断。

英文摘要:

We theoretically justify the recent empirical finding of [Teh et al., 2025] that a transformer pretrained on synthetically generated data achieves strong performance on empirical Bayes (EB) problems. We take an indirect approach to this question: rather than analyzing the model architecture or training dynamics, we ask why a pretrained Bayes estimator, trained under a prespecified training distribution, can adapt to arbitrary test distributions. Focusing on Poisson EB problems, we identify the existence of universal priors such that training under these priors yields a near-optimal regret bound of $\widetilde{O}(\frac{1}{n})$ uniformly over all test distributions. Our analysis leverages the classical phenomenon of posterior contraction in Bayesian statistics, showing that the pretrained transformer adapts to unknown test distributions precisely through posterior contraction. This perspective also explains the phenomenon of length generalization, in which the test sequence length exceeds the training length, as the model performs Bayesian inference using a generalized posterior.

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。

↑