表格深度学习论文趋势 2023–2026

主要结论

• 表格深度学习论文从 2023 年的 150 篇降到 2026 年的 33 篇，热度在回落。
• NeurIPS 是发表格深度学习论文最多的会议，2023–2026 年共 133 篇，占四个会议总量的 62%。
• 交叉方向里涨得最快的是驾驶仿真：同时属于两个主题的论文从 2023 年的 3 篇增至 2026 年的 6 篇。
• 在「图与结构化数据」的 5 个主题里，它的 2026 年论文数排第 5；同组增长最快的时间序列是 3.2 倍，表格深度学习是 0.2 倍。

代表论文

按年份和会议均匀挑选，方便快速看到每一届的样貌

• NeurIPS 2026Coherent Routing in Decision Trees: Phase-Interference Learning for Interpretable Tabular Prediction
• ICML 2026Density-Guided Robust Counterfactual Explanations on Tabular Data under Model Multiplicity
• ICLR 2026Harpoon: Generalised Manifold Guidance for Conditional Tabular Diffusion
• CVPR 2026MultiModalPFN Extending Prior-Data Fitted Networks for Multimodal Tabular Learning
• NeurIPS 2025ConTextTab: A Semantics-Aware Tabular In-Context Learner
• ICML 2025Latent Score-Based Reweighting for Robust Classification on Imbalanced Tabular Data
• ICLR 2025Revisiting Nearest Neighbor for Tabular Data: A Deep Tabular Baseline Two Decades Later
• NeurIPS 2024WikiDBs: A Large-Scale Corpus Of Relational Databases From Wikidata
• ICML 2024BiSHop: Bi-Directional Cellular Learning for Tabular Data with Generalized Sparse Modern Hopfield Model
• ICLR 2024GRANDE: Gradient-Based Decision Tree Ensembles for Tabular Data
• NeurIPS 2023Accurate Interpolation for Scattered Data through Hierarchical Residual Refinement
• ICML 2023Compressing Tabular Data via Latent Variable Estimation
• ICLR 2023TabCaps: A Capsule Neural Network for Tabular Data Classification with BoW Routing
• NeurIPS 2026LLM-Enhanced Random Forests in Orthogonal Hyperbolic Subspaces for Tabular Learning
• ICML 2026FedHPro: Federated Hyper-Prototype Learning via Gradient Matching
• NeurIPS 2025Feature-aware Modulation for Learning from Temporal Tabular Data
• ICML 2025TabFlex: Scaling Tabular Learning to Millions with Linear Attention
• ICLR 2025SAVA: Scalable Learning-Agnostic Data Valuation
• NeurIPS 2024Evaluating language models as risk scores
• ICML 2024Tabular Insights, Visual Impacts: Transferring Expertise from Tables to Images

全部表格深度学习论文

共 214 篇，第 1 / 5 页

1. CVPR 2026MultiModalPFN Extending Prior-Data Fitted Networks for Multimodal Tabular Learning
2. ICLR 2026Harpoon: Generalised Manifold Guidance for Conditional Tabular Diffusion
3. ICML 2026Attacks on Machine-Text Detectors Retain Stylistic Fingerprints
4. ICML 2026Cascaded Flow Matching for Heterogeneous Tabular Data with Mixed-Type Features
5. ICML 2026Density-Guided Robust Counterfactual Explanations on Tabular Data under Model Multiplicity
6. ICML 2026FedHPro: Federated Hyper-Prototype Learning via Gradient Matching
7. ICML 2026Float8@2bits: Entropy Coding Enables Data-Free Model Compression
8. ICML 2026GraphPFN: A Prior-Data Fitted Graph Foundation Model
9. ICML 2026HInT: Hypergraph Infusion at the Structural Layers Improves Table Understanding
10. ICML 2026Inverse Entropic Optimal Transport Solves Semi-supervised Learning via Data Likelihood Maximization
11. ICML 2026LassoFlexNet: a Flexible Neural Architecture for Tabular Data
12. ICML 2026OMP: One-step Meanflow Policy with Directional Alignment
13. ICML 2026On the Fragility of Data Attribution When Learning Is Distributed
14. ICML 2026pTNAS: Progressive Neural Architecture Search for Tabular Data
15. ICML 2026Removing Noise, not Finding Gold: Quality Filtering for Large-Scale Pretraining
16. ICML 2026TabPack: Efficient Hyperparameter Ensembles for Tabular Deep Learning
17. ICML 2026TabSwift: An Efficient Tabular Foundation Model with Row-Wise Attention
18. ICML 2026TAG: Tangential Amplifying Guidance for Hallucination-Resistant Sampling
19. ICML 2026The Data Manifold under the Microscope
20. ICML 2026TopBench: A Benchmark for Implicit Predictive Reasoning in Tabular Question Answering
21. ICML 2026Towards Efficient LLMs Annealing with Principled Sample Selection
22. ICML 2026TRIM: Token-wise Attention-Derived Saliency for Data-Efficient Instruction Tuning
23. ICML 2026Tuning-Free One-Class Discriminant Learning for Tabular Anomaly Detection
24. ICML 2026Unveiling the Role of Data Uncertainty in Tabular Deep Learning
25. NeurIPS 2026Coherent Routing in Decision Trees: Phase-Interference Learning for Interpretable Tabular Prediction
26. NeurIPS 2026Global linear convergence of entropy-regularized softmax policy gradient beyond tabular MDPs
27. NeurIPS 2026LLM-Enhanced Random Forests in Orthogonal Hyperbolic Subspaces for Tabular Learning
28. NeurIPS 2026STRABLE: Benchmarking Tabular Machine Learning with Strings
29. NeurIPS 2026Strengthening LLMs for Tabular Prediction with Structural Priors
30. NeurIPS 2026TabBioMed: A Large-Scale Benchmark for Biomedical Tabular Learning
31. NeurIPS 2026TabClustPFN: A Prior-Fitted Network for Tabular Data Clustering
32. NeurIPS 2026TabK: Amortized Bayesian Estimation of the Number of Clusters in Tabular Data
33. NeurIPS 2026When to Trust a PFN: Detecting Harmful Shift in Tabular Foundation Models
34. ICLR 2025AnoLLM: Large Language Models for Tabular Anomaly Detection
35. ICLR 2025Minimax Optimal Reinforcement Learning with Quasi-Optimism
36. ICLR 2025Revisiting Nearest Neighbor for Tabular Data: A Deep Tabular Baseline Two Decades Later
37. ICLR 2025SAVA: Scalable Learning-Agnostic Data Valuation
38. ICLR 2025TabM: Advancing tabular deep learning with parameter-efficient ensembling
39. ICLR 2025TabWak: A Watermark for Tabular Diffusion Models
40. ICML 2025Latent Score-Based Reweighting for Robust Classification on Imbalanced Tabular Data
41. ICML 2025TabFlex: Scaling Tabular Learning to Millions with Linear Attention
42. ICML 2025Understanding the Limits of Deep Tabular Methods with Temporal Shift
43. NeurIPS 2025CLIMB: Class-imbalanced Learning Benchmark on Tabular Data
44. NeurIPS 2025ConTextTab: A Semantics-Aware Tabular In-Context Learner
45. NeurIPS 2025Feature-aware Modulation for Learning from Temporal Tabular Data
46. NeurIPS 2025Sharp Gap-Dependent Variance-Aware Regret Bounds for Tabular MDPs
47. NeurIPS 2025TabArena: A Living Benchmark for Machine Learning on Tabular Data
48. ICLR 2024A Flexible Generative Model for Heterogeneous Tabular EHR with Missing Modality
49. ICLR 2024GRANDE: Gradient-Based Decision Tree Ensembles for Tabular Data
50. ICLR 2024Making Pre-trained Language Models Great on Tabular Prediction

常见问题

2023–2026 年四大会议一共有多少篇表格深度学习论文？

共 214 篇，按年份分别是 2023 年 150 篇、2024 年 17 篇、2025 年 14 篇、2026 年 33 篇。

NeurIPS 2026 有多少篇表格深度学习论文？

NeurIPS 2026 有 9 篇表格深度学习论文，上一届是 5 篇。

ICLR 2026 有多少篇表格深度学习论文？

ICLR 2026 有 1 篇表格深度学习论文，上一届是 6 篇。

ICML 2026 有多少篇表格深度学习论文？

ICML 2026 有 22 篇表格深度学习论文，上一届是 3 篇。

CVPR 2026 有多少篇表格深度学习论文？

CVPR 2026 有 1 篇表格深度学习论文。

哪个会议的表格深度学习论文最多？

NeurIPS，2023–2026 年共 133 篇。