Deep Learning Theory Retreat - 2025
Elma Hotel, Zikhron Yaakov - October 19-20, 2025
- Yotam Alexander (Nadav Cohen) - "Do Neural Networks Need Gradient Descent to Generalize? A Theoretical Study"
- Yuval Ran-Milo (Nadav Cohen) - "Transformers Can Provably Learn Topological Sorting"
- Yonatan Slutzky (Nadav Cohen) - "The implicit Bias of Structured State Space Models Can be Poisoned with Clean Labels"
- Sol Yarkoni (Roi Livni) - "Low Resource Reconstruction Attacks Through Benign Prompts"
- Liad Erez (Tomer Koren) - "Fast Rates in Bandit Multiclass Classification"
- Matan Schliserman (Tomer Koren) - "Multiclass Loss Geometry Matters for Generalization of Gradient Descent in Separable Classification"
- Shira Vansover-Hager (Tomer Koren) - "Rapid Overfitting of Multi-Pass SGD in Stochastic Convex Optimization"
- Nimrod Serok (Tal Pupko) - "Using Neural Networks to infer multiple sequence alignments"
- Ella Baumer (Tal Pupko) - "Adaptive Deep Reinforcement Learning for Phylogenetic Tree Reconstruction"
