information geometry - Robuta Search

https://openreview.net/forum?id=ApudXGKwIer&referrer=%5Bthe%20profile%20of%20Piotr%20A%20Sokol%5D(%2Fprofile%3Fid%3D~Piotr_A_Sokol2) Information Geometry of Orthogonal Initializations and Training | OpenReview Recently mean field theory has been successfully used to analyze properties of wide, random neural networks. It gave rise to a prescriptive theory for... information geometry orthogonal training openreview https://www.caltech.edu/campus-life-events/calendar/information-geometry-and-physics-seminar-17 Information, Geometry, and Physics Seminar - www.caltech.edu Toward a Control Theory of LLMs information geometry physics seminar www caltech https://www.caltech.edu/campus-life-events/calendar/information-geometry-and-physics-seminar-8 Information, Geometry, and Physics Seminar - www.caltech.edu Rigorous results about entropies in QFT information geometry physics seminar www caltech https://arxiv.org/abs/1207.6736 [1207.6736] Information geometry and sufficient statistics Abstract page for arXiv paper 1207.6736: Information geometry and sufficient statistics information geometry 1207 6736 sufficient statistics https://arxiv.org/abs/cond-mat/0401092 [cond-mat/0401092] Information Geometry and Phase Transitions Abstract page for arXiv paper cond-mat/0401092: Information Geometry and Phase Transitions information geometry cond phase transitions https://arxiv.org/abs/2502.00168 [2502.00168] Supervised Quadratic Feature Analysis: Information Geometry Approach for... Abstract page for arXiv paper 2502.00168: Supervised Quadratic Feature Analysis: Information Geometry Approach for Dimensionality Reduction information geometry 2502 00168 supervised quadratic https://www.preprints.org/manuscript/202408.2120 Cartan-Schouten Metrics for Information Geometry and Machine Learning[v1] | Preprints.org We study Cartan-Schouten metrics, explore invariant dual connections, and propose them as models for Information Geometry. Based on the underlying Riemannian... for information https://openreview.net/forum?id=rkg1ngrFPr Information Geometry of Orthogonal Initializations and Training | OpenReview nearly isometric DNN initializations imply low parameter space curvature, and a lower condition number, but that's not always great information geometry orthogonal training openreview https://www.caltech.edu/campus-life-events/calendar/information-geometry-and-physics-seminar-32 Information, Geometry, and Physics Seminar - www.caltech.edu Quantum Black Hole Entropy from a Localization Integral information geometry physics seminar www caltech https://www.caltech.edu/campus-life-events/calendar/information-geometry-and-physics-seminar-11 Information, Geometry, and Physics Seminar - www.caltech.edu Higher Information: The untold topological secrets of measures and states information geometry physics seminar www caltech https://openreview.net/forum?id=3wU2UX0voE The Information Geometry of Unsupervised Reinforcement Learning | OpenReview How can a reinforcement learning (RL) agent prepare to solve downstream tasks if those tasks are not known a priori? One approach is unsupervised skill... the information reinforcement learning geometry unsupervised openreview