undefined | Better HN

0 pointssrush4mo ago0 comments

Our primary focus is on RL post-training. We think that is the best way to get the model to be a strong interactive agent.

0 comments

comex4mo ago

So, yes, but you won’t say what the base model is? :)

It seems like a sort of sonnet model as a lot of people are reporting it like to spam documentation on Twitter like sonnet 4.5

j / k navigate · click thread line to collapse

0 pointssrush4mo ago0 comments

Our primary focus is on RL post-training. We think that is the best way to get the model to be a strong interactive agent.

comex4mo ago

So, yes, but you won’t say what the base model is? :)

It seems like a sort of sonnet model as a lot of people are reporting it like to spam documentation on Twitter like sonnet 4.5

j / k navigate · click thread line to collapse