Nathan Lévy @ICLR’25
@nathan_levyy
ML & Spatial Omics | PhD student @YosefLab | @ENS_ParisSaclay alum
I'm at #ICLR, presenting our work at #MLGenX & #LMRL workshops! - 2DE: a probabilistic method for differential expression across tissue niches - NOLAN: a self-supervised framework for mapping continuous tissue organization (don't miss @Kaiseki_Ou's oral on Monday 4:37pm!)
🚨ICML Paper Alert🚨 What if finding the right protein homologs wasn't a slow search, but a learned part of the model itself? We introduce 𝐏𝐫𝐨𝐭𝐫𝐢𝐞𝐯𝐞𝐫, an end-to-end framework that learns to retrieve the most useful homologs for self-supervised reconstruction! (1/12)
For our third Computational Spatial Biology seminar, we will speaking with @wencksternjo about VirTues, a foundation model framework for analyzing complex tissue architectures. Tune in this Wed, May 7 @ 11AM ET! Full schedule and last week's recording: csbseminar.github.io
See you tomorrow!
For our next Computational Spatial Biology seminar, we will talking with @SiyuHe7 about CORAL, a deep graph generative model learning integrated representations of multimodal spatial-omics data. Tune in next Wednesday, April 30 @ 12PM ET! Full schedule: csbseminar.github.io
To kick off the Computational Spatial Biology seminar series with @nathan_levyy, we will be discussing LUNA with @yistyu. LUNA is a generative AI model that reassembles complex tissue structures from gene expressions of cells. Tune in next Wednesday, April 9 @ 11AM ET!
Announcing the Computational Spatial Biology seminar! With @justjhong, we’re bringing together researchers at the intersection of AI & spatial biology to share their work. Kicking off April 9th with Yist Yu presenting LUNA! Explore the lineup here: csbseminar.github.io
So proud of my intern @Rotem_Shalita for her work at the #AiHub on deciphering T cell homing signatures for optimized CAR T cell engineering for neurodegeneration!
Hugely influential VAE paper, very well deserved test of time award
Excited to announce the release of scvi-tools v1.1 today! We’re including various new features, models, and critical bug fixes in this release, some of which I will highlight in this thread: docs.scvi-tools.org/en/1.1.0.post2…
Excited to share our study on how aging affects the human immune system at different tissues, with single cell genomics. Fantastic teamwork with @psimslab, Jo Jones, @DonnaFarber3 and @teichlab; and with Cambridge CBTM and @liveonny. Fresh on bioRxiv: doi.org/10.1101/2024.0…
Expanded Yosef Lab meeting at the SCG conference this week! In his talk Nir presented mrVI, a deep generative model for analyzing sample-level heterogeneity in single-cell RNA-seq datasets. Find out more: github.com/YosefLab/mrvi
Est-ce si étonnant que des chefs politiques du monde entier, aux idées opposées, en partagent une seule: planquer leur fric ?
Vallser: un tour à droite, un tour à gauche. Macroner: marcher de gauche à droite. Hollander: vallser en macronant. Pas facile!
Se sarkozir v. pron. Déf.: tout faire pour revenir en haut de l'affiche. Ex.: les efforts pathétiques de Sarah Bernhardt pour se sarkozir.
Je crains que dans les silences des hommes il n'y ait plus de ressentiments et de procès que de douceurs et d'éloges.
Les djihadistes ont réussi le tour de force de redonner de la légitimité aux paroles de la Marseillaise.
Tous ces veaux, vaches, moutons qui, après 9 jours de gloire à Paris, s'en retournent, fiers et déprimés, dans leurs étables anonymes.
Les petits rubans rouges accrochés à la poitrine des tyrans font de grosses taches rouges sur le front de nos présidents.
Les huîtres : ne nous ôtez pas notre accent circonflexe, c'est notre perle.