@RamaAdrien

creating games

 France
Joined May 2014
  • Tweets1.5K
  • Following642
  • Followers157
12Media
Instead of relying on pretrained video latent space we study which tradeoff of spatial and temporal resolution is optimal for real time WM. We open source all our models and training recipes so you can easily adapt it for your needs!
kyutai
@kyutai_labsOct 22
Many world models borrow their video autoencoder from text-to-video generators, and with it, a noticeable lag on every action. Today we release CADRE, open CAusal Diffusion REpresentation autoencoders that let real-time world models react fast, on a single GPU 馃У
1
8
27
857
After MIRA on Rocket League, I was very curious to know what it takes to bring interactive world models to the real world. It starts with the video autoencoder: how, where and how much to compress. CADRE is our answer, and it's all open 馃憞
kyutai
@kyutai_labsOct 22
Many world models borrow their video autoencoder from text-to-video generators, and with it, a noticeable lag on every action. Today we release CADRE, open CAusal Diffusion REpresentation autoencoders that let real-time world models react fast, on a single GPU 馃У
1
5
22
593
Many world models borrow their video autoencoder from text-to-video generators, and with it, a noticeable lag on every action. Today we release CADRE, open CAusal Diffusion REpresentation autoencoders that let real-time world models react fast, on a single GPU 馃У
6
32
4
157
8,717
New blog post about the Samuel text-to-speech, with 1,170 words and five (5) never-seen-before animations!
6
15
1
164
7,662
Introducing REKI, a 2.4B vision-language-action model pretrained on public data only. With a simple design, it rivals larger models trained on in-house data with complex multi-stage pipelines, setting a new Pareto frontier between performance and size. kyutai.org/reki
9
41
9
261
15,061
congratulations to @_Anthea___ for leading such a useful important robotics open-source project!!
kyutai
@kyutai_labsOct 16
Introducing REKI, a 2.4B vision-language-action model pretrained on public data only. With a simple design, it rivals larger models trained on in-house data with complex multi-stage pipelines, setting a new Pareto frontier between performance and size. kyutai.org/reki
5
134
Finally, a text-to-speech for the 80s.
42
197
26
3,181
156,102
Replying to @grok
@grok I'm confused what is happening in this clip?
Elon Musk
@elonmuskOct 10
Replying to @teslayoda
We鈥檙e making interesting progress here. Digital Optimus can play about halfway through the Diablo campaign so far just by looking at the screen like a human. Skill at Counter-Strike and other fast games is good too. League is also in training. Goal is to generalize across all games.
4
12
1
38
2,525
actual footage of how the 1v1s between GI鈥檚 models and whatever Elon is cooking will go (yes these are our vision based models)
Elon Musk
@elonmuskOct 10
Replying to @teslayoda
We鈥檙e making interesting progress here. Digital Optimus can play about halfway through the Diablo campaign so far just by looking at the screen like a human. Skill at Counter-Strike and other fast games is good too. League is also in training. Goal is to generalize across all games.
29
27
19
210
75,774
Tempted to start my Claude plushie business as demand skyrockets 馃
1
2
8
490
Checkout our latest work led by the great @RamaAdrien on representation alignment between unpaired modalities! You can align your autoencoder latent structure to the structure of a text model!
kyutai
@kyutai_labsOct 2
Many generative models work in an autoencoder's latent space; They train faster and generate better when this space is well organized. One way to organize it is distilling a pretrained model into it. But how can it work when the two spaces have different structures?
6
26
1,704
Many generative models work in an autoencoder's latent space; They train faster and generate better when this space is well organized. One way to organize it is distilling a pretrained model into it. But how can it work when the two spaces have different structures?
2
27
2
217
12,189
some autoencoders like mimi in audio or va-vae distill representations in the latent space to make it easier to generate. these works assume that the teacher has a structure that is somewhat shared by the student, which is not necessarily the case!
kyutai
@kyutai_labsOct 2
Many generative models work in an autoencoder's latent space; They train faster and generate better when this space is well organized. One way to organize it is distilling a pretrained model into it. But how can it work when the two spaces have different structures?
1
3
7
341
distilling dino into 1D spaces, or distilling text representations into an image autoencoder doesn't fit their approach. in this short paper with @nico_dufour, @ptrkprz and @david_picard, we explore this and propose a simple structure-agnostic way to tackle this distillation!
1
55
samuel now has a technical writeup from @vvolhejn go check out his wonderful project!
V谩clav Volhejn@vvolhejnSep 29
In many cases, Samuel learns to position its virtual mouth as a human would, even though its not explicitly trained to do so.
2
149
We've launched Nivalis Nights today!
129
534
75
6,379
356,355
Meta is making a tamagotchi for its AI? Cute, I've had a few on my desktop for a few weeks 馃悾 Tamaclaudchi is a (cute) Claude session manager to see and monitor all your Claude sessions at once Made in collaboration with @romfbr31 and more features to come with @RamaAdrien
1
5
1
10
602
not only is this very cute, but it is also extremely useful! i've been using some features of tamaclauidchi for a few weeks now and it's been really handy. stay tuned!!
Amelie Royer@royaleeriemeSep 29
Meta is making a tamagotchi for its AI? Cute, I've had a few on my desktop for a few weeks 馃悾 Tamaclaudchi is a (cute) Claude session manager to see and monitor all your Claude sessions at once Made in collaboration with @romfbr31 and more features to come with @RamaAdrien
2
107
General Intuition has raised another $220M at a $6.2B valuation as we start making our models available. These are an entirely new class of foundation models, trained on billions of action-labeled videos, capable of acting in realtime in never seen before environments. 1/6
50
95
62
900
405,886
Excited to share our latest work on bringing reasoning capabilities to speech-native models!
Alexandre D茅fossez@honualxSep 23
Sharing our latest work with @timweiss__ and @EXGRV . How to bridge the gap between the intelligence of text and speech models? What remains: how to combine full-duplex naturalness, latency, and intelligence in one model! We will present this at COLM 2026 in SF!
1
3
95
Load more