AIA historical bug in LLM scaling laws led the AI industry to train oversized, undertrained modelsGoogle DeepMind's Sander Dieleman highlighted the costly pre-Chinchilla compute error2 Sources, 94d ago, first seen 94d agoCombined views94.5K2 Sources, first seen 94d ago625 likesCombined views94.5K2 Sources, first seen 94d ago625 likes
