Learning Theory

Trustworthy Machine Learning through Data-Specific Indistinguishability ICML 2025 Truth Behind the Scene: Designing Evaluations Benchmarks to Assess LLMs’ Task-Specific Understanding over Test-Taking Strategies AAAI 2025 Tuning Algorithmic and Architectural Hyperparameters in Graph-Based Semi-Supervised Learning with Provable Guarantees UAI 2025 Two-Timescale Linear Stochastic Approximation: Constant Stepsizes Go a Long Way AISTATS 2025 Uncovering Scaling Laws for Large Language Models via Inverse Problems EMNLP 2025 Understanding Aggregations of Proper Learners in Multiclass Classification ALT 2025 Understanding Deep Representation Learning via Layerwise Feature Compression and Discrimination JMLR 2025 Understanding Expert Structures on Minimax Parameter Estimation in Contaminated Mixture of Experts AISTATS 2025 Understanding Generalization in Quantum Machine Learning with Margins ICML 2025 Understanding Model Ensemble in Transferable Adversarial Attack ICML 2025 Understanding Nonlinear Implicit Bias via Region Counts in Input Space ICML 2025 Understanding Sharpness Dynamics in NN Training with a Minimalist Example: The Effects of Dataset Difficulty, Depth, Stochasticity, and More ICML 2025 Understanding Stragglers in Large Model Training Using What-if Analysis OSDI 2025 Understanding the Forgetting of (Replay-based) Continual Learning via Feature Learning: Angle Matters ICML 2025 Understanding the Statistical Accuracy-Communication Trade-off in Personalized Federated Learning with Minimax Guarantees ICML 2025 Unified Analysis of Continuous Weak Features Learning with Applications to Learning from Missing Data ICML 2025 Universal Approximation Theorem of Deep Q-Networks ICML 2025 Universality of High-Dimensional Logistic Regression and a Novel CGMT under Dependence with Applications to Data Augmentation COLT 2025 Universal Rates for Multiclass Learning with Bandit Feedback COLT 2025 Universal Rates of ERM for Agnostic Learning COLT 2025 Unlocking Global Optimality in Bilevel Optimization: A Pilot Study ICLR 2025 Unlocking the Power of Rehearsal in Continual Learning: A Theoretical Perspective ICML 2025 Unraveling the Mechanics of Learning-Based Demonstration Selection for In-Context Learning ACL 2025 Unveiling Internal Reasoning Modes in LLMs: A Deep Dive into Latent Reasoning vs. Factual Shortcuts with Attribute Rate Ratio EMNLP 2025 User Behavior Prediction as a Generic, Robust, Scalable, and Low-Cost Evaluation Strategy for Estimating Generalization in LLMs ACL 2025