对比IMAGE
DeepSeek V4.1 Flash 1017 said to beat GPT-6 Astra on ML research tasks
bradthomasbrown says DeepSeek V4.1 Flash 1017 outperforms GPT-6 Astra on an ML research test where an agent builds a model to solve a benchmark instead of solving it directly.
Brad Brown
@bradthomasbrownArchived 11 days agoOriginally posted 2026.10.20
Brad Brown @bradthomasbrownOriginally posted 2026.10.20
DeepSeek-V4.1-Flash-1017 mogs Astra when it comes to ML research. Ex. take a benchmark, ask an agent to make a model of any variety that solves rather than solving it itself. Anthropic publicly admits sabotaging ML research. OpenAI too? Big fat fish wins by default? @teortaxesTex
x.com/bradthomasbrown/status/2112321533628858701Archived 2026.10.20 12:04