@BenjaminTrom
 Paris
Joined February 2022
  • Tweets179
  • Following182
  • Followers209
15Media
when will we have a jev but for next token prediction?!
1
51
Congrats to the team, amazing release!
Leonard Strouk
@_leostr_Sep 10
Millennium Prize Problems a making headlines, but your wheat 🌾 won't care ! Thrilled to introduce BOTANIC-1, our new plant DNA FM family. Outperform every model we tested on our benchmark. Weights: huggingface.co/collections/l… Paper: biorxiv.org/content/10.64898…

Botanic1 - a living-models Collection

Biological Omics Transformer for Agricultural and Nutritional trait Inference in Crops* Botanic1 is a family of plant gLMs by Living Models

huggingface.co
1
6
978
Benjamin Trom retweeted
Leonard Strouk
@_leostr_Sep 10
Millennium Prize Problems a making headlines, but your wheat 🌾 won't care ! Thrilled to introduce BOTANIC-1, our new plant DNA FM family. Outperform every model we tested on our benchmark. Weights: huggingface.co/collections/l… Paper: biorxiv.org/content/10.64898…

Botanic1 - a living-models Collection

Biological Omics Transformer for Agricultural and Nutritional trait Inference in Crops* Botanic1 is a family of plant gLMs by Living Models

huggingface.co
5
14
4
63
88,608
Mistral was an exceptional place to learn. I worked with some of the best people in the field at a pace I didn't think was possible, and I'm truly grateful.
1
7
414
Today Magic published its first result since I joined. Pretraining is the most closely guarded capability in AI, and the one frontier labs are measured against. Magic just showed it can do it more efficiently than anyone else and is the smallest team in the world training 1T llms
2
1
7
263
If you want high impact on frontier LLMs in a small team that works from first principles, I don't know a better place to be right now. And you'd get to work with me! Roles are in SF: magic.dev/careers magic.dev/blog/pretraining

Careers at Magic

We are a small team with a shared belief in the positive potential of responsibly deployed AGI. We value innate drive, creativity, and the ability to find clarity in uncharted domains.

magic.dev
1
203
Benjamin Trom retweeted
Magic
@magicailabsSep 8
Frontier pretraining is said to be a big-lab-only game. We don’t have 100k chips yet, so there’s only one way: algorithmic efficiency. Our new recipe matches DeepSeek V4 Pro’s pretrain using 50x less compute – that’s roughly half the FLOPs used for GPT3, or ~$0.5M on GB200. magic.dev/blog/pretraining
78
179
118
2,116
467,162
don't repeat this, but I've heard that @SchmidhuberAI actually invented hanzi ; before that chinese people were writing in pinyin
1
3
241
"generated with claude" is becoming the new "sent from my iPhone". technically true. completely useless. so i made the slop scale
1
1
7
220
more replies
so: self-assess the slop level. then put the audit result proudly in the pr body. I made a skill for that. side effect: doing the audit may force you to actually understand your own PR. github.com/limiteinductive/a…
1
2
102
@zomglings
1
70
y combinator should rebrand to GSpot at this point
2
154
i'm pretty sure that a claude codes outages has a measurable impact on birth rates
2
171
Load more