数据集与基准论文趋势 2023–2026
新数据集与评测协议
统计范围：NeurIPS、ICLR、ICML、CVPR 四个会议，2023–2026 年，共 5,856 篇被归入该主题的论文。
主要结论
- 数据集与基准论文从 2023 年的 937 篇增长到 2026 年的 1,991 篇，三年间扩大到约 2.1 倍。
- NeurIPS 是发数据集与基准论文最多的会议，2023–2026 年共 1,935 篇，占四个会议总量的 33%。
- 交叉方向里涨得最快的是AI 智能体：同时属于两个主题的论文从 2023 年的 8 篇增至 2026 年的 194 篇。
- 在「效率与系统」的 5 个主题里，它的 2026 年论文数排第 1；同组增长最快的硬件协同是 4.4 倍，数据集与基准是 2.1 倍。
各会议论文数
按年份和会议统计
| 年份 | NeurIPS | ICLR | ICML | CVPR | 合计 |
|---|---|---|---|---|---|
| 2023 | 227 | 162 | 166 | 382 | 937 |
| 2024 | 448 | 216 | 318 | 364 | 1,346 |
| 2025 | 637 | 419 | 189 | 337 | 1,582 |
| 2026 | 623 | 635 | 275 | 458 | 1,991 |
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 2026A Surrogate Perspective on Convergence of Fixed-Target DQN
- ICML 2026$\mu$pscaling small models: Principled warm starts and hyperparameter transfer
- ICLR 2026**TandemFoilSet**: Datasets for Flow Field Prediction of Tandem-Airfoil Through the Reuse of Single Airfoils
- CVPR 20264DEquine Disentangling Motion and Appearance for 4D Equine Reconstruction from
- NeurIPS 2025A Dataset for Distilling Knowledge Priors from Literature for Therapeutic Design
- ICML 2025Active Evaluation Acquisition for Efficient LLM Benchmarking
- ICLR 2025$F^3Set$: Towards Analyzing Fast, Frequent, and Fine-grained Events from Videos
- CVPR 20253D-GRAND: A Million-Scale Dataset for 3D-LLMs with Better Grounding and Less Hallucination
- NeurIPS 2024A Benchmark Dataset for Event-Guided Human Pose Estimation and Tracking in Extreme Conditions
- ICML 2024A connection between Tempering and Entropic Mirror Descent
- ICLR 2024A Benchmark Study on Calibration
- CVPR 2024Data Valuation and Detections in Federated Learning
- NeurIPS 2023$\mathbf{\mathbb{E}^{FWI}}$: Multiparameter Benchmark Datasets for Elastic Full Waveform Inversion of Geophysical Properties
- ICML 2023A Picture of the Space of Typical Learnable Tasks
- ICLR 2023$\Lambda$-DARTS: Mitigating Performance Collapse by Harmonizing Operation Selection among Cells
- CVPR 2023A Bag-of-Prototypes Representation for Dataset-Level Applications
- NeurIPS 2026A Theoretical Bridge Between Long-Tailed Recognition and Continual Learning
- ICML 2026A Semantically Consistent Dataset for Data-Efficient Query-Based Universal Sound Separation
- ICLR 2026A Fictional Q&A Dataset for Studying Memorization and Knowledge Acquisition
- CVPR 2026Advancing Cancer Prognosis with Hierarchical Fusion of Genomic Proteomic and
全部数据集与基准论文
共 1,935 篇，第 13 / 39 页
- NeurIPS 2026We Need to Improve Benchmarks in AI for Mathematics
- NeurIPS 2026WebSpatial: A Benchmarking Framework of Spatial Intelligence via Web-based Runtime Environments
- NeurIPS 2026Weird Generalization from Narrow Finetuning: Persona Shifts and Inductive Backdoors
- NeurIPS 2026Welfare, Improvability, and Variance: A Principal-Agent Approach to Optimal Benchmark Item Aggregation
- NeurIPS 2026What Cohort INRs Encode, and Where to Freeze Them
- NeurIPS 2026What Fits (Into Few Tokens) Doesn't Overfit: Compression and Generalization in ML Research Agents
- NeurIPS 2026What Sketches Tell Us about LVLMs: Conventions, Grounding, and Localisation
- NeurIPS 2026What Transformer FFNs Never See: Theory, Diagnosis, and Lightweight Remediation
- NeurIPS 2026When and Why Does Multi-Agent Debate Fail and Does It Really Underperform?
- NeurIPS 2026When and Why is Optimistic Multiplicative Weights Slow? The Geometry of Energy Dissipation
- NeurIPS 2026When Do Learned State Representations Break Sensitivity Analysis?
- NeurIPS 2026When Helpfulness Becomes Sycophancy: Sycophancy is a Boundary Failure Between Social Alignment and Epistemic Integrity in Large Language Models
- NeurIPS 2026When Sanitization Becomes the Trigger: Defense-Triggered Backdoor Attacks
- NeurIPS 2026Where Tabular Foundation Models Falter on Genetic Data: Datasets That Expose and Provide a Path to Address the Gap
- NeurIPS 2026Where to Connect? Boosting MLLMs via Dynamic Gated Pathways across ALL ViT and LLM Layers
- NeurIPS 2026Which Way Did It Move? Diagnosing and Overcoming Directional Motion Blindness in Video LLMs
- NeurIPS 2026Who&When Pro: Can LLMs Really Attribute Failures in AI Agents?
- NeurIPS 2026Why Heavy-Tailed Weights Predict Model Quality
- NeurIPS 2026WildBox: A Dataset and Benchmark for Aerial Monocular 3D Detection of African Savanna Wildlife
- NeurIPS 2026Winfree Oscillatory Neural Network
- NeurIPS 2026Workspace-Bench 1.0: Benchmarking AI Agents on Workspace Tasks with Large-Scale File Dependencies
- NeurIPS 2026X-Palm: Paired Multispectral-to-Smartphone Dataset for Cross-Domain Palmprint Authentication
- NeurIPS 2026XL-SafetyBench: A Country-Grounded Cross-Cultural Benchmark for LLM Safety and Cultural Sensitivity
- NeurIPS 20253D-RAD: A Comprehensive 3D Radiology Med-VQA Dataset with Multi-Temporal Analysis and Diverse Diagnostic Tasks
- NeurIPS 20253EED: Ground Everything Everywhere in 3D
- NeurIPS 20254D3R: Motion-Aware Neural Reconstruction and Rendering of Dynamic Scenes from Monocular Videos
- NeurIPS 2025A Closer Look at Graph Transformers: Cross-Aggregation and Beyond
- NeurIPS 2025A Closer Look at TabPFN v2: Understanding Its Strengths and Extending Its Capabilities
- NeurIPS 2025A Controllable Examination for Long-Context Language Models
- NeurIPS 2025A Dataset for Distilling Knowledge Priors from Literature for Therapeutic Design
- NeurIPS 2025A faster training algorithm for regression trees with linear leaves, and an analysis of its complexity
- NeurIPS 2025A Minimalist Example of Edge-of-Stability and Progressive Sharpening
- NeurIPS 2025A Multi-Task Benchmark for Abusive Language Detection in Low-Resource Settings
- NeurIPS 2025A Multimodal Benchmark for Framing of Oil & Gas Advertising and Potential Greenwashing Detection
- NeurIPS 2025A Set of Generalized Components to Achieve Effective Poison-only Clean-label Backdoor Attacks with Collaborative Sample Selection and Triggers
- NeurIPS 2025A Signed Graph Approach to Understanding and Mitigating Oversmoothing
- NeurIPS 2025A Standardized Benchmark for Multilabel Antimicrobial Peptide Classification
- NeurIPS 2025A Technical Report on “Erasing the Invisible”: The 2024 NeurIPS Competition on Stress Testing Image Watermarks
- NeurIPS 2025A Unifying View of Linear Function Approximation in Off-Policy RL Through Matrix Splitting and Preconditioning
- NeurIPS 2025Absence Bench: Language Models Can’t See What’s Missing
- NeurIPS 2025AbstentionBench: Reasoning LLMs Fail on Unanswerable Questions
- NeurIPS 2025Activation-Informed Merging of Large Language Models
- NeurIPS 2025Adapting to Stochastic and Adversarial Losses in Episodic MDPs with Aggregate Bandit Feedback
- NeurIPS 2025ADG: Ambient Diffusion-Guided Dataset Recovery for Corruption-Robust Offline Reinforcement Learning
- NeurIPS 2025Aeolus: A Multi-structural Flight Delay Dataset
- NeurIPS 2025AGC-Drive: A Large-Scale Dataset for Real-World Aerial-Ground Collaboration in Driving Scenarios
- NeurIPS 2025AgentDAM: Privacy Leakage Evaluation for Autonomous Web Agents
- NeurIPS 2025AGENTIF: Benchmarking Large Language Models Instruction Following Ability in Agentic Scenarios
- NeurIPS 2025AgentRecBench: Benchmarking LLM Agent-based Personalized Recommender Systems
- NeurIPS 2025Aggregation Hides Out-of-Distribution Generalization Failures from Spurious Correlations
在 Claude、Cursor 里直接问这个趋势
接上 Paperlist MCP 后，写综述、挑方向或准备投稿时，不用离开对话框就能拿到这页的数字：每年各会议的论文数、增长倍数、涨得最快的子方向和代表论文，AI 会按你的问题整理。
可以直接复制这句话问 AI
用 paperlist 查数据集与基准在 NeurIPS 2023–2026 的论文趋势，哪一年涨得最快？
AI 会调用的工具：get_topic_stats {"topic": "dataset-and-benchmarks"}
常见问题
- 2023–2026 年四大会议一共有多少篇数据集与基准论文？
- 共 5,856 篇，按年份分别是 2023 年 937 篇、2024 年 1,346 篇、2025 年 1,582 篇、2026 年 1,991 篇。
- NeurIPS 2026 有多少篇数据集与基准论文？
- NeurIPS 2026 有 623 篇数据集与基准论文，上一届是 637 篇。
- ICLR 2026 有多少篇数据集与基准论文？
- ICLR 2026 有 635 篇数据集与基准论文，上一届是 419 篇。
- ICML 2026 有多少篇数据集与基准论文？
- ICML 2026 有 275 篇数据集与基准论文，上一届是 189 篇。
- CVPR 2026 有多少篇数据集与基准论文？
- CVPR 2026 有 458 篇数据集与基准论文，上一届是 337 篇。
- 哪个会议的数据集与基准论文最多？
- NeurIPS，2023–2026 年共 1,935 篇。