Tensor Programs IVb: Adaptive Optimization in the Infinite-Width Limit von Yang, Greg, Littwin, Etai

Volltext bestellen

Tight conditions for when the NTK approximation is valid von Boix-Adsera, Enric, Littwin, Etai

Volltext bestellen

Enhancing JEPAs with Spatial Conditioning: Robust and Efficient Representation Learning von Littwin, Etai, Thilak, Vimal, Gopalakrishnan, Anand

Volltext bestellen

Tensor Programs IIb: Architectural Universality of Neural Tangent Kernel Training Dynamics von Yang, Greg, Littwin, Etai

Volltext bestellen

On the Convex Behavior of Deep Neural Networks in Relation to the Layers' Width von Littwin, Etai, Wolf, Lior

Volltext bestellen

UI-JEPA: Towards Active Perception of User Intent through Onscreen User Activity von Fu, Yicheng, Anantha, Raviteja, Vashisht, Prabal, Cheng, Jianpeng, Littwin, Etai

Volltext bestellen

Tensor Programs IVb: Adaptive Optimization in the Infinite-Width Limit von Yang, Greg, Littwin, Etai

Volltext

Regularizing by the Variance of the Activations' Sample-Variances von Littwin, Etai, Wolf, Lior

Volltext bestellen

Tight conditions for when the NTK approximation is valid von Boix-Adsera, Enric, Littwin, Etai

Volltext

How JEPA Avoids Noisy Features: The Implicit Bias of Deep Linear Self Distillation Networks von Littwin, Etai, Saremi, Omid, Advani, Madhu, Thilak, Vimal, Nakkiran, Preetum, Huang, Chen, Susskind, Joshua

Volltext bestellen

Enhancing JEPAs with Spatial Conditioning: Robust and Efficient Representation Learning von Littwin, Etai, Thilak, Vimal, Gopalakrishnan, Anand

Volltext

Tensor Programs IIb: Architectural Universality of Neural Tangent Kernel Training Dynamics von Yang, Greg, Littwin, Etai

Volltext

Vanishing Gradients in Reinforcement Finetuning of Language Models von Razin, Noam, Zhou, Hattie, Saremi, Omid, Thilak, Vimal, Bradley, Arwen, Nakkiran, Preetum, Susskind, Joshua, Littwin, Etai

Volltext bestellen

What Algorithms can Transformers Learn? A Study in Length Generalization von Zhou, Hattie, Bradley, Arwen, Littwin, Etai, Razin, Noam, Saremi, Omid, Susskind, Josh, Bengio, Samy, Nakkiran, Preetum

Volltext bestellen

Learning Representation from Neural Fisher Kernel with Low-rank Approximation von Zhang, Ruixiang, Zhai, Shuangfei, Littwin, Etai, Susskind, Josh

Volltext bestellen

When can transformers reason with abstract symbols? von Boix-Adsera, Enric, Saremi, Omid, Abbe, Emmanuel, Bengio, Samy, Littwin, Etai, Susskind, Joshua

Volltext bestellen

Transformers learn through gradual rank increase von Boix-Adsera, Enric, Littwin, Etai, Abbe, Emmanuel, Bengio, Samy, Susskind, Joshua

Volltext bestellen

The Loss Surface of Residual Networks: Ensembles and the Role of Batch Normalization von Littwin, Etai, Wolf, Lior

Volltext bestellen

On Random Kernels of Residual Architectures von Littwin, Etai, Galanti, Tomer, Wolf, Lior

Volltext bestellen

Stabilizing Transformer Training by Preventing Attention Entropy Collapse von Zhai, Shuangfei, Likhomanenko, Tatiana, Littwin, Etai, Busbridge, Dan, Ramapuram, Jason, Zhang, Yizhe, Gu, Jiatao, Susskind, Josh

Volltext bestellen