1
Heavy-Tailed Principal Component Analysis
重尾数据下PCA的经典方法失效？这篇论文提出了一种鲁棒主成分分析方法，突破二阶矩限制。
arXiv:2603.11308v3 Announce Type: replace Abstract: Principal Component Analysis (PCA) is a cornerstone of dimensionality reduction, yet its classical…
2
Statistical Early Stopping for Reasoning Models
提出用统计方法为推理模型设计早停策略，显著提升计算效率而不损失准确度。
arXiv:2602.13935v2 Announce Type: replace-cross Abstract: While LLMs have seen substantial improvement in reasoning capabilities, they also sometimes …
3
Quantifying Ranking Uncertainty in LLM Benchmarks
量化LLM基准测试中排名的不确定性，为模型评估提供统计严谨性。
arXiv:2607.16259v1 Announce Type: new Abstract: Pretrained models are typically ranked on multi-task leaderboards to assess their effectiveness across…
4
Ensuring Trustworthy Online A/B Testing: Addressing Five Key Questions on CUPED
系统解答CUPED在A/B测试中的五个关键问题，帮你避免实验数据分析陷阱。
arXiv:2606.18750v1 Announce Type: cross Abstract: A/B testing has become the gold standard for data-driven decision-making in large-scale online exper…
5
From Uncertain Judgments to Calibrated Rankings: Conformal Elo Estimation for LLM Evaluation
用共形预测校准LLM评估中的Elo排名，解决判分偏差与不可传递性问题。
arXiv:2606.13221v2 Announce Type: replace Abstract: Evaluating new large language models typically requires costly human annotation campaigns at scale…
6
PERRY: Policy Evaluation with Confidence Intervals using Auxiliary Data
一种利用辅助数据构建置信区间的新策略评估方法，显著提升统计推断效率与稳健性。
arXiv:2507.20068v2 Announce Type: replace Abstract: Off-policy evaluation (OPE) methods estimate the value of a new reinforcement learning (RL) policy…
7
Beyond Averages: Evaluating LLMs on Human Survey Replication at the Distributional Level
从均值到分布：论文提出评估LLM复现人类调查结果的新方法，更精准衡量AI模拟真实人群的能力。
arXiv:2606.09013v1 Announce Type: new Abstract: LLMs are increasingly used to simulate human survey responses, but prior work has mainly evaluated rep…
8
Smooth Multi-Policy Causal Effect Estimation in Longitudinal Settings
纵向数据下多策略因果效应平滑估计，提升精度与可解释性，适合因果推断研究者。
arXiv:2605.14284v2 Announce Type: replace Abstract: Comparative evaluation of multiple dynamic treatment policies is essential for healthcare and poli…
9
The Illusion of Intervention: Your LLM-Simulated Experiment is an Observational Study
大语言模型模拟实验只是观察性研究，干预错觉被无情拆穿
arXiv:2605.20767v1 Announce Type: cross Abstract: Large language models (LLMs) show potential as simulators of human behavior, offering a scalable way…
10
TEA-Time: Transporting Effects Across Time
因果推断新方法：跨时间运输效应，助力时间序列分析更精准
arXiv:2603.07018v2 Announce Type: replace-cross Abstract: Treatment effects estimated from a randomized controlled trial are local not only to the stu…