Theory · learning theory papers, 2023–2026

Key findings

• Theory · learning theory went from 86 papers in 2023 to 482 in 2026, roughly 5.6x in three years.
• NeurIPS publishes the most Theory · learning theory work: 540 papers from 2023 to 2026, 40% of the four-venue total.
• The fastest-growing overlap is with Social aspects · accountability transparency and interpretability: papers tagged with both rose from 0 in 2023 to 33 in 2026.
• Among the 5 topics under Theory and Statistics, it ranks #2 by 2026 paper count; the fastest grower in the group, Theory · optimization, is up 4.7x versus 5.6x here.

Representative papers

Spread across years and venues so every edition shows up

• NeurIPS 2026A dynamical systems theory of reward-modulated learning in linear recurrent networks
• ICML 2026A Fine-Grained Understanding of Uniform Convergence for Halfspaces
• ICLR 2026A Law of Data Reconstruction for Random Features (And Beyond)
• NeurIPS 2025A Geometric Analysis of PCA
• ICML 2025A Generalization Theory for Zero-Shot Prediction
• ICLR 2025A Solvable Attention for Neural Scaling Laws
• CVPR 2025Dynamic Stereotype Theory Induced Micro-expression Recognition with Oriented Deformation
• NeurIPS 20244+3 Phases of Compute-Optimal Neural Scaling Laws
• ICML 2024A Statistical Theory of Regularization-Based Continual Learning
• ICLR 2024A Sublinear Adversarial Training Algorithm
• CVPR 2024Accurate Training Data for Occupancy Map Prediction in Automated Driving Using Evidence Theory
• NeurIPS 2023$\varepsilon$-fractional core stability in Hedonic Games.
• ICML 2023Robust Collaborative Learning with Linear Gradient Overhead
• ICLR 2023Bispectral Neural Networks
• NeurIPS 2026A Mechanistic Investigation of Theory of Mind in a Large Language Model
• ICML 2026A Random Matrix Theory of Masked Self-Supervised Learning
• ICLR 2026A New Approach to Controlling Linear Dynamical Systems
• NeurIPS 2025Approximation theory for 1-Lipschitz ResNets
• ICML 2025A Two-Stage Learning-to-Defer Approach for Multi-Task Learning
• ICLR 2025A Theoretical Analysis of Self-Supervised Learning for Vision Transformers

All Theory · learning theory papers

147 papers, page 1 of 3

1. ICLR 2025A Solvable Attention for Neural Scaling Laws
2. ICLR 2025A Theoretical Analysis of Self-Supervised Learning for Vision Transformers
3. ICLR 2025A Theoretical Perspective: How to Prevent Model Collapse in Self-consuming Training Loops
4. ICLR 2025A Theory of Initialisation's Impact on Specialisation
5. ICLR 2025A Unified Theory of Quantum Neural Network Loss Landscapes
6. ICLR 2025Adapting Multi-modal Large Language Model to Concept Drift From Pre-training Onwards
7. ICLR 2025Adversarial Training Can Provably Improve Robustness: Theoretical Analysis of Feature Learning Process Under Structured Data
8. ICLR 2025Algorithmic Stability Based Generalization Bounds for Adversarial Training
9. ICLR 2025Almost Optimal Batch-Regret Tradeoff for Batch Linear Contextual Bandits
10. ICLR 2025An Online Learning Theory of Trading-Volume Maximization
11. ICLR 2025Analyzing Neural Scaling Laws in Two-Layer Networks with Power-Law Data Spectra
12. ICLR 2025Ask, and it shall be given: On the Turing completeness of prompting
13. ICLR 2025Associative memory and dead neurons
14. ICLR 2025Asymptotic Analysis of Two-Layer Neural Networks after One Gradient Step under Gaussian Mixtures Data with Structure
15. ICLR 2025Attention layers provably solve single-location regression
16. ICLR 2025Bandit Learning in Matching Markets with Indifference
17. ICLR 2025Bayesian Treatment of the Spectrum of the Empirical Kernel in (Sub)Linear-Width Neural Networks
18. ICLR 2025Benign Overfitting in Out-of-Distribution Generalization of Linear Models
19. ICLR 2025Binary Losses for Density Ratio Estimation
20. ICLR 2025Block Verification Accelerates Speculative Decoding
21. ICLR 2025Bounds on $L_p$ Errors in Density Ratio Estimation via $f$-Divergence Loss Functions
22. ICLR 2025Bridging the Gap Between f-divergences and Bayes Hilbert Spaces
23. ICLR 2025Broadening Target Distributions for Accelerated Diffusion Models via a Novel Analysis Approach
24. ICLR 2025Can Neural Networks Achieve Optimal Computational-statistical Tradeoff? An Analysis on Single-Index Model
25. ICLR 2025Can One Modality Model Synergize Training of Other Modality Models?
26. ICLR 2025Chain-of-Thought Provably Enables Learning the (Otherwise) Unlearnable
27. ICLR 2025Classic but Everlasting: Traditional Gradient-Based Algorithms Converge Fast Even in Time-Varying Multi-Player Games
28. ICLR 2025Computational Explorations of Total Variation Distance
29. ICLR 2025Compute-Optimal LLMs Provably Generalize Better with Scale
30. ICLR 2025Conditional Diffusion Models are Minimax-Optimal and Manifold-Adaptive for Conditional Distribution Estimation
31. ICLR 2025Connectome Mapping: Shape-Memory Network via Interpretation of Contextual Semantic Information
32. ICLR 2025Correcting the Mythos of KL-Regularization: Direct Alignment without Overoptimization via Chi-Squared Preference Optimization
33. ICLR 2025Decentralized Sporadic Federated Learning: A Unified Algorithmic Framework with Convergence Guarantees
34. ICLR 2025Decomposition Polyhedra of Piecewise Linear Functions
35. ICLR 2025Demystifying Online Clustering of Bandits: Enhanced Exploration Under Stochastic and Smoothed Adversarial Contexts
36. ICLR 2025Diffusion Transformer Captures Spatial-Temporal Dependencies: A Theory for Gaussian Process Data
37. ICLR 2025Distribution-Specific Agnostic Conditional Classification With Halfspaces
38. ICLR 2025Divergence of Neural Tangent Kernel in Classification Problems
39. ICLR 2025Dynamic Assortment Selection and Pricing with Censored Preference Feedback
40. ICLR 2025Exact Community Recovery under Side Information: Optimality of Spectral Algorithms
41. ICLR 2025Exploratory Preference Optimization: Harnessing Implicit Q*-Approximation for Sample-Efficient RLHF
42. ICLR 2025Exploring The Loss Landscape Of Regularized Neural Networks Via Convex Duality
43. ICLR 2025Extending Mercer's expansion to indefinite and asymmetric kernels
44. ICLR 2025Feature Averaging: An Implicit Bias of Gradient Descent Leading to Non-Robustness in Neural Networks
45. ICLR 2025Flavors of Margin: Implicit Bias of Steepest Descent in Homogeneous Neural Networks
46. ICLR 2025From Lazy to Rich: Exact Learning Dynamics in Deep Linear Networks
47. ICLR 2025From Sparse Dependence to Sparse Attention: Unveiling How Chain-of-Thought Enhances Transformer Sample Efficiency
48. ICLR 2025Fundamental Limitations on Subquadratic Alternatives to Transformers
49. ICLR 2025Generalizability of Neural Networks Minimizing Empirical Risk Based on Expressive Power
50. ICLR 2025Generalization Bounds and Model Complexity for Kolmogorov–Arnold Networks

FAQ

How many Theory · learning theory papers were published at NeurIPS, ICLR, ICML and CVPR from 2023 to 2026?

1,350 in total: 86 in 2023, 393 in 2024, 389 in 2025, 482 in 2026.

How many Theory · learning theory papers are at NeurIPS 2026?

NeurIPS 2026 has 112 Theory · learning theory papers, versus 106 the year before.

How many Theory · learning theory papers are at ICLR 2026?

ICLR 2026 has 200 Theory · learning theory papers, versus 147 the year before.

How many Theory · learning theory papers are at ICML 2026?

ICML 2026 has 170 Theory · learning theory papers, versus 132 the year before.

How many Theory · learning theory papers are at CVPR 2025?

CVPR 2025 has 4 Theory · learning theory papers, versus 1 the year before.

Which conference publishes the most Theory · learning theory papers?

NeurIPS, with 540 papers from 2023 to 2026.