←back to thread

LLaVA-O1: Let Vision Language Models Reason Step-by-Step

(arxiv.org)
176 points lnyan | 1 comments | 18 Nov 24 09:44 UTC | HN request time: 0.001s | source
Show context
startupsfail ◴[18 Nov 24 17:36 UTC] No.42174762[source]▶
Generating data with OpenAI model AND copying the approach from OpenAI model. This is a bit unsatisfactory, its like saying you wrote some working code, while in fact you’ve decompiled the binary and then compiled it again.
replies(1): >>42175872 #
exe34 ◴[18 Nov 24 19:19 UTC] No.42175872[source]▶
well if you have working code at the end, you made progress. closedAI can pull any model at any time for a profit.
replies(1): >>42200516 #
1. startupsfail ◴[21 Nov 24 02:35 UTC] No.42200516[source]▶
Yes, agreed. And it’s not like OpenAI isn’t doing the same thing, in a sense. Data was originally sampled from human annotations.