Datasets and Benchmarks papers, 2023–2026

Key findings

• Datasets and Benchmarks went from 937 papers in 2023 to 1,991 in 2026, roughly 2.1x in three years.
• NeurIPS publishes the most Datasets and Benchmarks work: 1,935 papers from 2023 to 2026, 33% of the four-venue total.
• The fastest-growing overlap is with AI Agents: papers tagged with both rose from 8 in 2023 to 194 in 2026.
• Among the 5 topics under Efficiency and Systems, it ranks #1 by 2026 paper count; the fastest grower in the group, Hardware-Aware ML, is up 4.4x versus 2.1x here.

Representative papers

Spread across years and venues so every edition shows up

• 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

All Datasets and Benchmarks papers

623 papers, page 13 of 13

601. NeurIPS 2026We Need to Improve Benchmarks in AI for Mathematics
602. NeurIPS 2026WebSpatial: A Benchmarking Framework of Spatial Intelligence via Web-based Runtime Environments
603. NeurIPS 2026Weird Generalization from Narrow Finetuning: Persona Shifts and Inductive Backdoors
604. NeurIPS 2026Welfare, Improvability, and Variance: A Principal-Agent Approach to Optimal Benchmark Item Aggregation
605. NeurIPS 2026What Cohort INRs Encode, and Where to Freeze Them
606. NeurIPS 2026What Fits (Into Few Tokens) Doesn't Overfit: Compression and Generalization in ML Research Agents
607. NeurIPS 2026What Sketches Tell Us about LVLMs: Conventions, Grounding, and Localisation
608. NeurIPS 2026What Transformer FFNs Never See: Theory, Diagnosis, and Lightweight Remediation
609. NeurIPS 2026When and Why Does Multi-Agent Debate Fail and Does It Really Underperform?
610. NeurIPS 2026When and Why is Optimistic Multiplicative Weights Slow? The Geometry of Energy Dissipation
611. NeurIPS 2026When Do Learned State Representations Break Sensitivity Analysis?
612. NeurIPS 2026When Helpfulness Becomes Sycophancy: Sycophancy is a Boundary Failure Between Social Alignment and Epistemic Integrity in Large Language Models
613. NeurIPS 2026When Sanitization Becomes the Trigger: Defense-Triggered Backdoor Attacks
614. NeurIPS 2026Where Tabular Foundation Models Falter on Genetic Data: Datasets That Expose and Provide a Path to Address the Gap
615. NeurIPS 2026Where to Connect? Boosting MLLMs via Dynamic Gated Pathways across ALL ViT and LLM Layers
616. NeurIPS 2026Which Way Did It Move? Diagnosing and Overcoming Directional Motion Blindness in Video LLMs
617. NeurIPS 2026Who&When Pro: Can LLMs Really Attribute Failures in AI Agents?
618. NeurIPS 2026Why Heavy-Tailed Weights Predict Model Quality
619. NeurIPS 2026WildBox: A Dataset and Benchmark for Aerial Monocular 3D Detection of African Savanna Wildlife
620. NeurIPS 2026Winfree Oscillatory Neural Network
621. NeurIPS 2026Workspace-Bench 1.0: Benchmarking AI Agents on Workspace Tasks with Large-Scale File Dependencies
622. NeurIPS 2026X-Palm: Paired Multispectral-to-Smartphone Dataset for Cross-Domain Palmprint Authentication
623. NeurIPS 2026XL-SafetyBench: A Country-Grounded Cross-Cultural Benchmark for LLM Safety and Cultural Sensitivity

FAQ

How many Datasets and Benchmarks papers were published at NeurIPS, ICLR, ICML and CVPR from 2023 to 2026?

5,856 in total: 937 in 2023, 1,346 in 2024, 1,582 in 2025, 1,991 in 2026.

How many Datasets and Benchmarks papers are at NeurIPS 2026?

NeurIPS 2026 has 623 Datasets and Benchmarks papers, versus 637 the year before.

How many Datasets and Benchmarks papers are at ICLR 2026?

ICLR 2026 has 635 Datasets and Benchmarks papers, versus 419 the year before.

How many Datasets and Benchmarks papers are at ICML 2026?

ICML 2026 has 275 Datasets and Benchmarks papers, versus 189 the year before.

How many Datasets and Benchmarks papers are at CVPR 2026?

CVPR 2026 has 458 Datasets and Benchmarks papers, versus 337 the year before.

Which conference publishes the most Datasets and Benchmarks papers?

NeurIPS, with 1,935 papers from 2023 to 2026.