Can Stationary Distributions of Scale-Invariant Neural Networks Be Described by the Thermodynamics of an Ideal Gas?
Paper presentation, 35th International Joint Conference on Artificial Intelligence, Bremen, Germany
Language: English
Talks and presentations
To Stay or Not to Stay in the Pre-train Basin: Insights on Ensembling in Transfer Learning
Paper presentation, Fall into ML 2023: Conference on Machine Learning at HSE University, Moscow, Russia
Language: English
Self-supervised Pre-training with Masked Image Modeling
Research seminar, Bayesgroup, online
Language: Russian
On the Memorization Properties of Contrastive Learning
Poster session, Workshop on Overparameterization: Pitfalls & Opportunities at ICML, online
Language: English