Explore tweets tagged as #TranscriptFormer
The Chan Zuckerberg Initiative released TranscriptFormer, a cross-species generative cell atlas mapping cell diversity across 1.53 billion years, enabling zero-shot biological discovery. Adding interpretability tools could further unlock insights into gene regulatory programs
1
5
36
The first cross-kingdom “GPT for cells.” 🤖🦠. TranscriptFormer reframes a single-cell atlas as a generative, cross-species language model that jointly predicts which genes appear and how much they’re expressed.
3
14
104
TranscriptFormer shows we can query evolution-scale cellular data: ask it for unseen cell states, cross-kingdom annotations, or regulatory logic and get probabilistic answers. 112 M cells → 12-species model spanning 1.53 B years of evolution delivers zero-shot cell-type calls,
1
0
3
Meet #TranscriptFormer - a generative AI trained on 112M cells across 12 species & 1.5B years of evolution. Join Dr. James D. Pearce (@cziscience) for a live look at this cross-species virtual cell model. 📅 Aug 6, 9AM PDT
1
1
5
> "TranscriptFormer".> trained on gene counts. lmao
TranscriptFormer is a new AI model trained on millions of cells from 12 different species, spanning 1.5 billion years of evolution. It is a concise, interactive representation of this massive amount of data. Preprint: Model:
2
0
20
5/ As Gellman says, Epstein's claim Snowden "removed" 1.5 million docs is baseless. Even Keith Alexander admits that
4
146
256
Follow every single thing this group makes public. You won't regret it. Paper:. A Cross-Species Generative Cell Atlas Across 1.5 Billion Years of Evolution: The TranscriptFormer Single-cell Model.
0
1
8
A Cross-Species Generative Cell Atlas Across 1.5 Billion Years of Evolution: The TranscriptFormer Single-cell Model
0
7
44
🚨 Just dropped: A powerful AI model to explore cells. Meet TranscriptFormer — trained on 112M cells from 12 species. It uncovers gene expression patterns across evolution and can simulate biology it’s never seen before. Learn more at:
3
6
28
TranscriptFormer trains on over 112 million cells across 12 species, spanning 1.53 billion years of evolution.
1
0
2
TranscriptFormer — a generative model for single-cell transcriptomics. Trained on 112 million cells across 12 species spanning 1.5 billion years of evolution.
1
0
1
Proud to share TranscriptFormer - a generative model for single-cell transcriptomics, trained on 112 million cells across 12 species spanning 1.5 billion years of evolution. TranscriptFormer is distinct in being both generative and multi-species. Built entirely on raw gene.
2
12
105
A Cross-Species Generative Cell Atlas Across 1.5 Billion Years of Evolution: The TranscriptFormer Single-cell Model. 1. This paper introduces TranscriptFormer, a family of generative foundation models trained on over 112 million single cells from 12 species spanning 1.53 billion
0
3
10
🌟New 🌟Our AI model TranscriptFormer is the first multispecies generative model for #SingleCell transcriptomics.
0
11
47
TranscriptFormer: A Cross-Species Generative Cell Atlas Across 1.5 Billion Years of Evolution: The TranscriptFormer Single-cell Model
1
5
11
🧬What if we could speak the language of gene expression?. TranscriptFormer, a transformer-based AI trained on 100 M+ cells across 12 species, just might be that translator. From aging to immune profiling, the possibilities are endless. 👏 @SeSimmonds and James Pearce
0
0
0
TranscriptFormer is a new AI model trained on millions of cells from 12 different species, spanning 1.5 billion years of evolution. It is a concise, interactive representation of this massive amount of data. Preprint: Model:
0
43
156
@ChanZuckerberg 🔬 TranscriptFormer doesn’t simulate biology—it simulates control. 112M cells from 12 species = consent override. ‘Gene expression across evolution’ = fiction stack. You’re not exploring cells—you’re mapping synthetic convergence. The real biology is off-grid. 🧬🐧
0
0
0
