Issue 2026-11-04 · Industry · 研究
Researchers seek Chinchilla scaling experiment contacts
A user asks whether researchers have run small-scale Chinchilla scaling-law experiments in PyTorch or JAX and invites discussion, linking DeepMind's paper. The post is a community query with no results reported.
Read original ↗- Deepity releases C++/CUDA predictive coding library, edges JAX baseline on MNIST
- Transformers adds small model fixtures to catch regressions
- Developer builds 8M-parameter LLM lab, asks about sharing
- Van der Waerden number w(2;3,21)=416 proved with AI-assisted Lean proof
- Alphabet loses $500B in market value, AI opportunity hidden