Back to researchers
Johannes Welbl
Compute-optimal scaling for LLM training
Co-authored the Chinchilla paper—an anchor for how to spend compute when training LLMs.
Highlights
DeepMindChinchillaScalingCompute-optimal
Focus: Compute-optimal scaling for LLM training
Why it matters: Co-authored the Chinchilla paper—an anchor for how to spend compute when training LLMs.
Research Areas
DeepMindChinchillaScalingCompute-optimal