3 ms·
> No one from the list (also add Anthropic to that list) is SOTA right now I don't use OpenAI's products. Is there an area where they are clearly ahead? I thou
by xnx 2y ago
> No one from the list (also add Anthropic to that list) is SOTA right now
I don't use OpenAI's products. Is there an area where they are clearly ahead? I thought it was at least a tie in all categories.
- mrbungie 2y agoLast time I checked, o1 did much better in PlanBench that models from other providers indicating better perf in that area [1][2]. Probably that lead will rapidly decay as other players implement similar approaches to "reasoning"/LRMs. [1] https://www.arxiv.org/abs/2409.13373 https://www.arxiv.org/abs/2409.13373 [2] https://x.com/omarsar0/status/1838353480672563581 https://x.com/omarsar0/status/1838353480672563581