NeurIPS 2025 Explorer
Concepts
Authors
Glossary
johnsanterre.github.io
semantic anchors
3 papers
REPA Works Until It Doesn’t: Early-Stopped, Holistic Alignment Supercharges Diffusion Training
SEGA: Shaping Semantic Geometry for Robust Hashing under Noisy Supervision
XIFBench: Evaluating Large Language Models on Multilingual Instruction Following