Tabular data is the primary data format in industrial relational databases, underpinning modern data analytics and decision-making. However, the increasing scale of tabular data poses significant computational and storage challenges to learning-based analytical systems. This highlights the need for data-efficient learning, which enables effective model training and generalization using substantially fewer samples. Dataset condensation (DC) has emerged as a promising data-centric paradigm that synthesizes small yet informative datasets to preserve data utility while reducing storage and training costs. However, existing DC methods are computationally intensive due to reliance on complex gradient-based optimization. Moreover, they often overlook key characteristics of tabular data, such as heterogeneous features and class imbalance. To address these limitations, we introduce C$^{2}$TC (Class-Adaptive Clustering for Tabular Condensation), the first training-free tabular dataset condensation framework that jointly optimizes class allocation and feature representation, enabling efficient and scalable condensation. Specifically, we reformulate the dataset condensation objective into a novel class-adaptive cluster allocation problem (CCAP), which eliminates costly training and integrates adaptive label allocation to handle class imbalance. To solve the NP-hard CCAP, we develop HFILS, a heuristic local search that alternates between soft allocation and class-wise clustering to efficiently obtain high-quality solutions. Moreover, a hybrid categorical feature encoding (HCFE) is proposed for semantics-preserving clustering of heterogeneous discrete attributes. Extensive experiments on 10 real-world datasets demonstrate that C$^{2}$TC improves efficiency by at least 2 orders of magnitude over state-of-the-art baselines, while achieving superior downstream performance.
Chunked TabPFN: Exact Training-Free In-Context Learning for Long-Context Tabular Data
Department of Statistics, Texas A\&M University, College Station, TX（德克萨斯A&M大学统计学系）；Department of Electrical and Computer Engineering, University of Minnesota, Twin City, MN（明尼苏达大学电气与计算机工程系）
LLM Empowered Prototype Learning for Zero and Few-Shot Tasks on Tabular Data
School of Artificial Intelligence, Jilin University（吉林大学人工智能学院）；Engineering Research Center of Knowledge-Driven Human-Machine Intelligence, Ministry of Education, China（教育部知识驱动人机智能工程研究中心）；International Center of Future Science, Jilin University, China（吉林大学未来科学国际中心）；College of Computer Science and Software Engineering, Shenzhen University（深圳大学计算机科学与软件工程学院）；CSIRO’s Data61 and Monash University（CSIRO的数据61与墨尔本大学）
Towards Benchmarking Foundation Models for Tabular Data With Text
Department of Computer Science, University of Freiburg, Freiburg, Germany（弗赖堡大学计算机科学系）；Prior Labs, Freiburg, Germany（Prior实验室）；ELLIS Institute, Tubingen, Germany（图宾根ELLIS研究所）
ICLAD: In-Context Learning for Unified Tabular Anomaly Detection Across Supervision Regimes
Department of Electrical and Computer Engineering, McGill University, Canada（麦吉尔大学电气与计算机工程系，加拿大）；Mila - Quebec Artificial Intelligence Institute, Montreal, Canada（魁北克人工智能研究所，蒙特利尔，加拿大）
CTG-KrEW: Generating Synthetic Structured Contextually Correlated Content by Conditional Tabular GAN with K-Means Clustering and Efficient Word Embedding
Indian Institute of Technology Kharagpur（印度理工学院克勒格布尔分校）；Missouri University of Science and Technology（密苏里科技大学）
Class-aware and Augmentation-free Contrastive Learning from Label Proportion
The Hong Kong University of Science and Technology（香港科技大学）；The Hong Kong University of Science and Technology (Guangzhou)（香港科技大学（广州））；Tencent America（腾讯美国公司）；Tencent Europe B.V.（腾讯欧洲有限责任公司）
Bonsai: Gradient-free Graph Condensation for Node Classification
Yardi School of Artificial Intelligence（亚迪人工智能学院）；Department of Computer Science（计算机科学系）；Department of Chemical Engineering（化学工程系）；Indian Institute of Technology Delhi（德里印度理工学院）
Horizon-Free Reinforcement Learning in Polynomial Time: the Power of Stationary Policies
Tsinghua University（清华大学）；University of Washington（华盛顿大学）
本文给出了表马尔可夫决策过程（MDP）的第一个具有后悔界\emph{independent on the planning horizon}的多项式时间算法。具体来说，我们考虑表格MDP与$S$状态，$A$行动，规划地平线$H$，总奖励由$1$界定，和代理播放$K$情节。我们设计了一个算法，实现了$O\left(\mathrm{poly}(S,A,\log K)\sqrt{K}\right)$遗憾，与现有的界限，要么有一个额外的$\mathrm{polylog}(H)$依赖性~ \citep{zhang2020reinforcement}或有一个指数依赖于$S$ ~ \citep{li2021settling}。我们的结果依赖于一系列新的结构引理建立的近似功率，稳定性和浓度属性的平稳政策，这可以在其他问题中的应用马尔可夫链。
Avoiding Structural Failure Modes in Tabular Fair SSL: Online Primal-Dual Allocation under Confidence Gating
College of Computer Science and Technology, Jilin University, China（吉林大学计算机科学与技术学院）；Key Laboratory of Symbolic Computation and Knowledge Engineering of the Ministry of Education, Jilin University, China（教育部符号计算与知识工程重点实验室）
TabSOM: A tabular-to-image encoding method based on self-organizing maps
Rey Juan Carlos University (URJC)（雷胡安·卡洛斯大学）；Universidad Francisco de Vitoria（弗朗西斯科·德·维多利亚大学）；University of Castilla-La Mancha（卡斯蒂利亚-拉曼恰大学）；Yachay Tech University（亚奇理工大学）
Self-Improving Tabular Language Models via Iterative Reward-Guided Post-Training
Department of Engineering, University of Cambridge（剑桥大学工程系）；CISPA Helmholtz Center for Information Security, Saarbrücken, Germany（德国萨尔布吕肯信息安全中心）；The Alan Turing Institute, London（伦敦阿兰·图灵研究所）
