Deep Cogito
@DeepCogito
Followers
3K
Following
6
Media
0
Statuses
6
Love how @DeepCogito is building its models using self-play - turns out this makes models better both in reasoning and non-thinking modes. Kudos @drishanarora and @DeepCogito team!
It is intuitively easy to understand why self play *can* work for LLMs, if we are able to provide a value function at intermediate steps (although not as clearly guaranteed as in two-player zero-sum games). In chess / go / poker, we have a reward associated with every next
3
1
10
It is intuitively easy to understand why self play *can* work for LLMs, if we are able to provide a value function at intermediate steps (although not as clearly guaranteed as in two-player zero-sum games). In chess / go / poker, we have a reward associated with every next
Self play works so well in chess, go, and poker because those games are two-player zero-sum. That simplifies a lot of problems. The real world is messier, which is why we haven’t seen many successes from self play in LLMs yet. Btw @karpathy did great and I mostly agree with him!
12
10
37
Cogito 671B is an impressive model, a material improvement over DSV3… often better than Sonnet & 4o. Nicely done @drishanarora & @DeepCogito team!
A small update - we had more traffic than anticipated. However, the endpoints are now scalable on Together AI for all models, including the 671B MoE. Test out the model here: https://t.co/Od1NXYVBxU (A huge thanks to the folks at @togethercompute for making this happen so
1
2
23
Deep Cogito goes big, releasing 4 new open source hybrid reasoning models with self-improving 'intuition'
venturebeat.com
Arora explains this as a difference between searching for a path versus already knowing roughly where the destination lies.
2
8
20
We are putting out four frontier intelligence models today, each hitting heavy in its weight class. The large ones are among the smartest open models in the world right now. It has become increasingly clear that superintelligence is within reach. As we continue to iterate and
Today, we are releasing 4 hybrid reasoning models of sizes 70B, 109B MoE, 405B, 671B MoE under open license. These are some of the strongest LLMs in the world, and serve as a proof of concept for a novel AI paradigm - iterative self-improvement (AI systems improving themselves).
3
4
44
Today, we are releasing 4 hybrid reasoning models of sizes 70B, 109B MoE, 405B, 671B MoE under open license. These are some of the strongest LLMs in the world, and serve as a proof of concept for a novel AI paradigm - iterative self-improvement (AI systems improving themselves).
45
262
2K