Computational Foundations of Reliable Learning

More workshops at Annual Conference on Learning Theory 2026

• Learning in an Agentic World — Examines learning-theory foundations and open challenges for autonomous, agentic AI systems.
• Bridging Computational and Statistical Theories of ML — Connects computational and statistical perspectives on the theory of machine learning.
• Learning-Augmented Algorithms — Studies algorithms that use learned predictions to improve performance while retaining theoretical guarantees.
• Foundations of Learning Reasoning Models — Develops theoretical foundations for training and understanding models that perform reasoning.
• Equivalences Between Lower Bound Frameworks — Explores connections between lower-bound frameworks including low-degree polynomials and statistical physics predictions.
• Foundations of Post-training — Examines the theoretical foundations of post-training methods for large models.
• Algorithmic Landscape of Sampling in Generative Models — Surveys algorithms and theory for sampling in modern generative models.