Twitxy Saved

samsja

Verified account Prime Intellect

@samsja19

leading research at @PrimeIntellect
sf github.com/samsja Joined March 2020
2.7K Following 9.3K Followers 5.6K Posts

Latest posts 5

Pinned post

samsja Verified account @samsja19
INTELLECT-3 is our first model I can use daily

It's build using our open source stack by scaling RL of MoE over 512 H200 and pushing the sota at its size

Incredible proud of leading such a team of talented, dedicated and hard working individual collaborating together to push the open frontier

Open source super intelligence is coming, the pretrain are also cooking
Prime Intellect @PrimeIntellect Nov 27, 2025
Introducing INTELLECT-3: Scaling RL to a 100B+ MoE model on our end-to-end stack

Achieving state-of-the-art performance for its size across math, code and reasoning

Built using the same tools we put in your hands, from environments & evals, RL frameworks, sandboxes & more
View quoted post on X
28 replies 18 reposts 444 likes 59K views View on X
samsja Verified account @samsja19
Very excited about this release

RL is really amazing, even more with multi agent system
Max Rumpf Verified account @maxrumpf 6h
Introducing SID-2, a new search model built on compute-scaling swarms.

Our swarms find better results than Opus and Astra, while being 8x faster and 20x lower cost.

By 2030, AI will write an internet's worth of text every 24 hours. Current search pipelines were built on hand-designed rules, adding more won't scale to meet this incoming volume of data.

SID-2 is available on our API today. Try it on your most demanding searches.
View quoted post on X
samsja Verified account @samsja19
I will be speaking next week at infercon about inference for RL !
Olivier Pomel @oliveur 5d
We're running INFER/CON next week in SF

Think putting AI into prod. Getting inference to work. Performance. Costs. Long-running agents... Hard tech rather than high-level commentary.

The line-up is stellar, so I expect this to be quite good :-)
View quoted post on X
0 replies 4 reposts 68 likes 2.9K views View on X
samsja Verified account @samsja19
I am a big fan of cognition post training team, this is a very solid release
Cognition @cognition Nov 4
Introducing SWE-2 Flash, our new fast & efficient coding model.

On leading evals, it outperforms all American open-weight models, even ones 3x larger.

To balance transparency and cyber risks, we are building a program to securely share the weights with partners & researchers.
View quoted post on X
1 replies 0 reposts 23 likes 1.2K views View on X
samsja Verified account @samsja19
this is a great release, great model and great tech report, congrats to thinking machines team !
Thinking Machines Verified account @thinkymachines Nov 5
Introducing Inkling 2 Small. Scoring 31 on the AA Intelligence Index, it's a big jump over its predecessor.

Inkling 2 Small is available for fine-tuning and inference on Tinker today, with full weight release planned soon.

thinkingmachines.ai/news/inkling-2…
View quoted post on X