Daniel Kokotajlo comments on [Link] Training Compute-Optimal Large Language Models

[Daniel Kokotajlo]

OK, good to know. I look forward to seeing the performance trends updated with the new scaling paradigm/​law.