Most HuggingFace Daily Papers aren't for me. So here's one specifically for the scientific machine learning community.
Every day, it scans the latest arXiv and OpenReview submissions, keeps only the papers relevant to the methods space — operator learning, physics-informed ML, neural ODEs, PDE solvers, and the applied math of neural networks — and delivers each one with a two-sentence AI-generated summary: what problem it tackles, what method it introduces.
Basically everything that I personally care about.
Operator Learning PDE Foundation Models Physics-Informed ML
Generative Models for Simulation Differentiable Simulation
ML-Accelerated Solvers Equation Discovery & Dynamical Systems
LLMs for Scientific Computing UQ & Inverse Problems
Mathematical Analysis of Neural Networks
Browse by date — a date strip at the top. Latest day selected by default. Every past day archived.
Filter by subfield — narrow the feed to exactly the topics you follow.
Search — find any paper by title or author across the full archive.
Flip for abstract — click any card to flip it and read the full abstract, with LaTeX rendered.