#machine learning
9 articles
Latest
Seven New Papers Show AI Cracking Materials, Physics, and Gene Problems
Deep learning advances inverse design, predicts fluid dynamics, infers gene networks, and discovers symbolic equations from raw data.
Train-to-Test Scaling Rebalances LLM Compute Budgets
New framework optimizes total AI costs by accounting for inference-time scaling alongside training.
NVIDIA Builds Fast Multilingual OCR with Synthetic Data
New Nemotron OCR v2 model processes text across languages without massive labeled datasets.
New PyTorch Repo Teaches Distributed Training From Scratch
Developer releases educational codebase implementing five parallelism strategies without high-level abstractions.
WSU Researchers Deploy AI Spectral Imaging for Plastic Sorting
Spectral imaging combined with machine learning improves automated identification of recyclable plastic types.
Netflix Releases VOID: Video Object Deletion Model
Netflix open-sources VOID, a video inpainting model that removes objects and their physical interactions from footage.
ICML 2026 Review Policy Debate: 100 Responses Split on Scoring
Community feedback reveals Policy B may score higher, but Policy A demonstrates stronger reviewer confidence levels.
Hugging Face Open Source Ecosystem Expands With Storage Tools
New storage buckets feature and record model uploads reshape how developers build with open models.
Hugging Face Open Source Ecosystem Hits New Scale in Spring 2026
Major infrastructure upgrades and storage tools reshape how researchers share machine learning models.