Great use of Botanic 1!
So happy to highlight one of the many applications of Botanic-1, our latest plant FM, on @googlegemma 's website !
teapawt.org/GoogleAI/status/210715…
Pierre Manceron retweeted
So happy to highlight one of the many applications of Botanic-1, our latest plant FM, on @googlegemma 's website !
teapawt.org/GoogleAI/status/210715…
Pierre Manceron retweeted
🌱🌱🌱 a HUGE step towards designing to food of tomorrow -- from the DNA up 🧬🧬🧬
🙏 BOTANIC-1 just dropped, a family of models to do just that (beating all prev models)
All open on @huggingface 🤗 + more useful than Millennium Problems
More in 🧵
Pierre Manceron retweeted
Ok kind of hard to squeeze in between Navier-Stokes BSD and Hodge ... but I'll try anyway ; )
We are super proud to release what we think is the strongest FM on plant genomics so far BOTANIC-1 🌱
Presenting our latest paper, accepted in the ECCV MEDFMB workshop: RadMatch
A new way to evaluate radiology report generation : being auditable, it provides feedback on important errors
arxiv.org/abs/2609.01470
Curia 2 has been accepted at @MICCAI_Society !
Antoine Saporta, Lead ML Researcher at @raidium_med will present it in an oral session
We'll also host a event, dm if interested 👀
Paper: arxiv.org/abs/2604.01987
Pierre Manceron retweeted
🤔 I noticed that the head of a GPT-like model can be applied to its intermediate layers to predict the next token. We can even determine where to stop inference threshold on the maximum logits. Is this a widely-known phenomenon? Could this approach speed up inference ? (1/4)
Pierre Manceron retweeted
Natural language has GPT3, Programming has OpenAI Codex, DALL-E has CLIP: Radiology needs a foundation model, to empower the radiologist with a new generation of AI assistance.
A short read about foundation models
medium.com/raidium/is-ai-for…
#foundationmodel #radiology #gpt3
Pierre Manceron retweeted
Owkin listed for the second year in a row in @CBinsights #AI100 list of #AI startups redefining industries! 🎉 It's awesome to be included among so many global key players. #medicalresearch #AIforhealthcare #GoOwkin
cbinsights.com/research/arti…
Pierre Manceron retweeted
We love seeing all the progress on differential privacy & ML, but we were frustrated that there aren’t great efficient implementations of the application of DP to DCNNs; Ouch! We proposed today some techniques that can speed things up a bit in PyTorch: arxiv.org/abs/1912.06015
Pierre Manceron retweeted
Another great moment for Owkin Lab! 👏Our Radiology Team wins again at JFR, this time applying their machine learning prowess to Sarcopenia screening. Great Work!🎉
Gagnant du data challenge sarcopenie #jfr2019 @OWKINscience. Félicitations !
















