Comment on DeepSeek: Inference-Time Scaling for Generalist Reward ModelingparentComments−gmerc1yIf it was Elon is even more stupid than he lets on becauseDS3: 5M training run Grok3: 400M training runfor 2% difference in the benchmarks.−resters1yThey probably pulled the plug at the last minute to switch to DeepSeek.
Comments
If it was Elon is even more stupid than he lets on because
DS3: 5M training run Grok3: 400M training run
for 2% difference in the benchmarks.
They probably pulled the plug at the last minute to switch to DeepSeek.