24 Fields Medal winners have written a letter of objection

Traditional analyses of gradient descent show that when the largest eigenvalue of the Hessian, also known as the sharpness $S(θ)$, is bounded by $2/η$, training is "stable" and the training loss decre...

Extension for Visual Studio Code - Adds LaTeX math rendering to Claude Code chat — inline $...$, display $$...$$, and \(...\) / \[...\]

We introduce Convex Distance Operator Transport (CDOT), the first convex optimal transport framework that aligns distributions across heterogeneous domains by jointly preserving feature correspondence...

An ICML 2026 tutorial on making probabilistic predictions trustworthy for downstream decision-making and collaboration.

A theoretical and empirical study of why placing momentum before orthogonalization makes Muon work, framed as a spectral filter on the gradient stream.

Muon has recently demonstrated strong empirical performance in large language model training, but the theoretical role of momentum in Muon remains unclear. Existing analyses of Muon either remove mome...

This page provides the information regarding the open call for applications for FY2025. Applications only be accepted from Japanese organizations.