@trislaz

Researcher @Microsoft Research, PhD @Mines_Paris. Alumni @ENS_ULM. Computer vision - histopathology.

 Paris, France
trislaz.github.io
Joined April 2011
  • Tweets55
  • Following681
  • Followers74
1Media
Biology doesn't operate in silos, and neither should the AI representation of it. Quine is an early-stage research effort to create a multimodal world model of biology. By connecting insights across biological scales and modalities, Quine helps scientists computationally search a space far larger than intuition allows and prioritize hypotheses before they reach the lab. Experimental results provide important feedback, helping researchers sharpen future research directions. msft.it/6014acaRn
16
36
20
199
41,970
We will soon be hiring postdocs, engineers and researchers (on site in Heidelberg, DE) in: - Biomedical AI (multimodal foundation models, biomarkers, agents) - Robotics (lab automation, physical action surfaces for agents) - AI social sciences (multi-agent systems, agent "civilizations") - Deploying AI in a large hospital system More info at imai-heidelberg.de and kather.ai If you are interested in learning more, reach out to office@kather.ai

Institute for Medical AI | Heidelberg

Building safe, clinically useful artificial intelligence for medicine.

imai-heidelberg.de
7
62
334
18,962
As many other oncologists will also attest, we were taught this was a dead end It was, dogmatically, never going to work — kras was too much of a “greasy ball” to be targeted And yet here we are, with truly meaningful survival curves👇 Inspiration on multiple levels #ASCO26
Dr. Catharine Young
@DrCatharineYMay 31
What I love about this news is that this is not simply a new pancreatic cancer drug. It’s that KRAS, one of cancer biology’s most notorious undruggable targets, has finally become druggable. A powerful reminder that persistence in science can eventually unlock the impossible.
17
169
10
1,085
91,229
Trislaz retweeted
Ryan Peters
@ryanirlMay 18
Biological networks too :) Here is the neural geometry of mice navigating a figure-8 maze.
Goodfire
@GoodfireAIMay 7
Neural networks might speak English, but they think in shapes. Understanding their rich *neural geometry* is key to understanding how they work – and to debugging and controlling them with precision. Starting today, we’re releasing a series of posts on this research agenda. 🧵
17
80
7
630
66,177
My course on consciousness is now online: six 90´ lectures featuring both history and recent experimental results on ignition, P3b, working memory, neural manifolds… unpublished results too. tinyurl.com/cz9yb4z3
21
148
5
633
78,100
Wow..!
Daniel Bear
@recursus16 Jul 2024
We at @noetik_ai are thrilled to introduce the *first world model of spatial, multimodal cancer biology*: Oncology Counterfactual Therapeutics Oracle (OCTO) Technical report below + 🧶 on how we trained OCTO directly on patient data and use it to find new cancer drug targets!
188
The boundary between trainable and untrainable neural network hyperparameter configurations is *fractal*! And beautiful! Here is a grid search over a different pair of hyperparameters -- this time learning rate and the mean of the parameter initialization distribution.
26
167
63
1,593
251,068
Still blows my mind that you can take ANY MIT COURSE you want, online, for free. Machine learning? Nuclear physics? Quantum computing? All there. Truly unreal.
180
1,576
274
12,698
6,060,517
An honor to be external PhD examiner for #AI pathology superstar Tristan Lazard, supervised by @ThomasW04410315 at @Mines_Paris @psl_univ. Love the ambiente 🇪🇺🇫🇷🍷 & the discussions about self-supervised learning, foundation models & biomarkers 👉linkinghub.elsevier.com/retr…
1
2
29
2,065
V1.0! Free phone-formatted pdf: fleuret.org/public/lbdl.pdf Pocket book: fleuret.org/lbdl-order Note that given an other recent prototype everything should be okay with the pocket book version, but I have not *yet* checked this one (it takes ~2 weeks to get it)
7
73
8
357
36,798
This is absolutely bananas. There is NOTHING like this on the instrumental record. Its likely impact is probably immeasurable. But is it leading our news bulletins & on newspaper front pages? Nope.
217
2,738
346
7,297
1,392,522
Seeing so many empty posters & missing authors at #CVPR2023 is heartbreaking - how many? 20%? Many PhD students worked hard but this absurd visa system jeopardized their chance to proudly present their work. I know that PCs @CVPR took action but this was largely insufficient… 1/
2
18
4
176
35,778
I am looking for a postdoc in Computer Vision. We have several exciting projects in the field of Digital Pathology, where we aim at predicting treatment response from Whole Slide Images (WSI). WSI are huge and complex, containing hundreds of thousands of individual cells.
1
10
14
3,763
🥇🥉🥉Elle l’a fait ! Après avoir terminé cinq fois sur la deuxième marche du podium en coupe du monde, Oriane Bertone remporte sa première victoire à Prague. Flavy Cohaut et Mejdi Schalck décrochent le bronze ! 👉 bit.ly/3oRhvrF 📷 Jan Virt - IFSC
1
24
3
182
15,625
ÅNGSTRÖM-RESOLUTION FLUORESCENCE MICROSCOPY. Can we do structural biology with an optical microscope? Yes! Check out how we resolved *single bases of DNA* and much more using fluorescence microscopy, out today in @Nature! nature.com/articles/s41586-0… @JungmannLab @MPI_Biochem

Ångström-resolution fluorescence microscopy

Nature - The authors introduce a single-molecule DNA-barcoding method, resolution enhancement by sequential imaging, that improves the resolution of fluorescence microscopy down to the...

nature.com
29
371
75
1,298
181,355
🙂Au Collège de France, les #enseignements sont #gratuits & libres d’accès. 🙂Certains le savent, d’autres l’ignorent. 🙂Que ceux qui le savent partagent ce tweet avec ceux qui l’ignorent 😉 college-de-france.fr
42
2,678
87
3,945
644,927
Very happy to share our latest preprint 🎉 "Fully transformer-based biomarker prediction from #colorectal #cancer histology". This is a massive collaborative study in >9000 patients which is led by @sophiajwagner: arxiv.org/abs/2301.09617 #AI (1/4)
3
28
3
137
33,938
Shoutout to the legend who answers literally every single PyTorch question out there!
54
191
42
3,888
645,808
Happy to share our latest publication in @NatureMedicine 🩺🧬🔬 nature.com/articles/s41591-0…. We designed a multistain deep learning model and trained, validated, and tested it on imaging data of different immune cell subtypes of over 1000 patients with #colorectal #cancer. Our …
12
49
3
215
82,049
Reinforcement learning from human feedback (RLHF) can teach LLMs a variety of interesting skills. As an example, Sparrow, a chatbot developed by @DeepMind, is taught (via RLHF) to support its factual claims by finding relevant information on Google... 🧵 [1/7]
3
74
4
529
105,572
Load more