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

389 papers, page 7 of 8

301. NeurIPS 2025Certifying Stability of Reinforcement Learning Policies using Generalized Lyapunov Functions
302. NeurIPS 2025Computable universal online learning
303. NeurIPS 2025Computational Efficiency under Covariate Shift in Kernel Ridge Regression
304. NeurIPS 2025Conformal Inference under High-Dimensional Covariate Shifts via Likelihood-Ratio Regularization
305. NeurIPS 2025Constant Bit-size Transformers Are Turing Complete
306. NeurIPS 2025Constrained Linear Thompson Sampling
307. NeurIPS 2025Contextual Dynamic Pricing with Heterogeneous Buyers
308. NeurIPS 2025Curl Descent : Non-Gradient Learning Dynamics with Sign-Diverse Plasticity
309. NeurIPS 2025Cyclic Counterfactuals under Shift–Scale Interventions
310. NeurIPS 2025Disentangling Hyperedges through the Lens of Category Theory
311. NeurIPS 2025Distributional Training Data Attribution: What do Influence Functions Sample?
312. NeurIPS 2025Do Neural Networks Need Gradient Descent to Generalize? A Theoretical Study
313. NeurIPS 2025Eluder dimension: localise it!
314. NeurIPS 2025Equivariance by Contrast: Identifiable Equivariant Embeddings from Unlabeled Finite Group Actions
315. NeurIPS 2025Exact Expressive Power of Transformers with Padding
316. NeurIPS 2025Exploiting Vocabulary Frequency Imbalance in Language Model Pre-training
317. NeurIPS 2025Fast attention mechanisms: a tale of parallelism
318. NeurIPS 2025Gaussian Process Upper Confidence Bound Achieves Nearly-Optimal Regret in Noise-Free Gaussian Process Bandits
319. NeurIPS 2025How Particle System Theory Enhances Hypergraph Message Passing
320. NeurIPS 2025Identifiability of Deep Polynomial Neural Networks
321. NeurIPS 2025Improved Balanced Classification with Theoretically Grounded Loss Functions
322. NeurIPS 2025Information-Theoretic Discrete Diffusion
323. NeurIPS 2025Infrequent Exploration in Linear Bandits
324. NeurIPS 2025Is Grokking a Computational Glass Relaxation?
325. NeurIPS 2025Knee-Deep in C-RASP: A Transformer Depth Hierarchy
326. NeurIPS 2025Last Iterate Convergence in Monotone Mean Field Games
327. NeurIPS 2025Learning from Interval Targets
328. NeurIPS 2025Learning Interestingness in Automated Mathematical Theory Formation
329. NeurIPS 2025Learning Linear Attention in Polynomial Time
330. NeurIPS 2025Learning Theory for Kernel Bilevel Optimization
331. NeurIPS 2025Length Generalization via Auxiliary Tasks
332. NeurIPS 2025Limitations of Normalization in Attention
333. NeurIPS 2025Low Precision Streaming PCA
334. NeurIPS 2025Machine Unlearning under Overparameterization
335. NeurIPS 2025Marginal-Nonuniform PAC Learnability
336. NeurIPS 2025Metric Automata Theory: A Unifying Theory of RNNs
337. NeurIPS 2025MindForge: Empowering Embodied Agents with Theory of Mind for Lifelong Cultural Learning
338. NeurIPS 2025Multidimensional Bayesian Utility Maximization: Tight Approximations to Welfare
339. NeurIPS 2025Neural Entropy
340. NeurIPS 2025Non-Asymptotic Analysis Of Data Augmentation For Precision Matrix Estimation
341. NeurIPS 2025Non-Stationary Lipschitz Bandits
342. NeurIPS 2025Non-Uniform Multiclass Learning with Bandit Feedback
343. NeurIPS 2025Nyström-Accelerated Primal LS-SVMs: Breaking the $O(an^3)$ Complexity Bottleneck for Scalable ODEs Learning
344. NeurIPS 2025On Agnostic PAC Learning in the Small Error Regime
345. NeurIPS 2025On Linear Mode Connectivity of Mixture-of-Experts Architectures
346. NeurIPS 2025On Minimax Estimation of Parameters in Softmax-Contaminated Mixture of Experts
347. NeurIPS 2025On the Emergence of Linear Analogies in Word Embeddings
348. NeurIPS 2025On the Optimality of the Median-of-Means Estimator under Adversarial Contamination
349. NeurIPS 2025On the Surprising Effectiveness of Large Learning Rates under Standard Width Scaling
350. NeurIPS 2025On the VC dimension of deep group convolutional neural 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.