Nils Trost
@trostnils.bsky.social
33 followers 37 following 4 posts
Posts Media Videos Starter Packs
Pinned
trostnils.bsky.social
I'm very excited to finally share the main work of my PhD!
We explored the evolutionary dynamics of gene regulation and expression during gonad development in primates. We cover among others: X chromosome dynamics (incl. in a developing XXY testis), gene regulatory networks and cell type evolution.
kaessmannlab.bsky.social
We are delighted to share our new preprint “The evolution of gene regulatory programs controlling gonadal development in primates” www.biorxiv.org/content/10.1...
Reposted by Nils Trost
aicoffeebreak.bsky.social
Ever wondered how Energy-Based Models (EBMs) work and how they differ from normal neural networks?
☕️ We go over EBMs and then dive into the Energy-Based Transformers paper to make LLMs that refine guesses, self-verify, and could adapt compute to problem difficulty.
Reposted by Nils Trost
aicoffeebreak.bsky.social
The world’s largest NLP conference with almost 2,000 papers presented, ACL 2025 just took place in Vienna! 🎓✨ Here is a quick snapshot of the event via a short interview with one of the authors whose work caught my attention.
🎥 Watch: youtu.be/GBISWggsQOA
Reposted by Nils Trost
kaessmannlab.bsky.social
👇 !!!
louisjeantetfdn.bsky.social
Register to attend the #LouisJeantetSymposium on vertebrate genome #evolution. Organised by Svante Pääbo and
@kaessmannlab.bsky.social . Free event, full line up of speakers 👉 www.jeantet.ch/en/

📆Tuesday 14 October, 08h15
📍CMU - Auditoire Alex-F. Müller (A250)
@genevunige.bsky.social
trostnils.bsky.social
Grob über den Daumen gepeilt hat also jeder Steuerzahler im Durschschnitt 2€ bezahlt um 0.000003875 Menschen an der Grenze zurückzuweisen.
Oder: ca 258.000€ pro zurückgewiesenem Menschen.
Reposted by Nils Trost
aicoffeebreak.bsky.social
How do LLMs pick the next word? They don’t choose words directly: they only output word probabilities. 📊 Greedy decoding, top-k, top-p, min-p are methods that turn these probabilities into actual text.
Reposted by Nils Trost
aicoffeebreak.bsky.social
Excited to be at ACL 2025 in Vienna this week 🇦🇹 #ACL2025
I’m always up for a chat about reasoning models, NLE faithfulness, synthetic data generation, or the joys and challenges of explaining AI on YouTube.

If you're around, let’s connect!
Reposted by Nils Trost
aicoffeebreak.bsky.social
🤖 Can we trust AI in science?
I'm excited to be speaking at the final event of the Young Marsilius Fellows 2025, themed "Dancing with Right & Wrong?" – a title that feels increasingly relevant these days.
I'll be joining a panel on "(How) can we trust AI in science?" to discuss questions like:
trostnils.bsky.social
I want to take this opportunity to thank everyone at the @kaessmannlab.bsky.social (including the past members) for creating such a welcoming, stimulating and fun environment!
Special thanks to @amir-f.bsky.social, the shared first-author, who performed the majority of the sequencing experiments.
trostnils.bsky.social
I'm very excited to finally share the main work of my PhD!
We explored the evolutionary dynamics of gene regulation and expression during gonad development in primates. We cover among others: X chromosome dynamics (incl. in a developing XXY testis), gene regulatory networks and cell type evolution.
kaessmannlab.bsky.social
We are delighted to share our new preprint “The evolution of gene regulatory programs controlling gonadal development in primates” www.biorxiv.org/content/10.1...
Reposted by Nils Trost
aicoffeebreak.bsky.social
Excited to share that I’ll be joining the Summer School “AI and Human Values” this September at the Marsilius-Kolleg of Heidelberg University as a speaker. I'll be giving an introduction to how large language models actually work—before the summer school dives deeper into broader implications.
trostnils.bsky.social
To give the print head of the the teleprinter enough time to reach the beginning of the new line!
Reposted by Nils Trost
aicoffeebreak.bsky.social
Long videos are a nightmare for language models—too many tokens, slow inference. ☠️
We explain STORM ⛈️, a new architecture that improves long video LLMs using Mamba layers and token compression. Reaches better accuracy than GPT-4o on benchmarks and up to 8× more efficiency.

📺 youtu.be/uMk3VN4S8TQ
Token-Efficient Long Video Understanding for Multimodal LLMs | Paper explained
YouTube video by AI Coffee Break with Letitia
youtu.be
Reposted by Nils Trost
kaessmannlab.bsky.social
A great article highlighting the broader implications of the three parallel recent papers from our group and our wonderful colleagues in Spain and Belgium...
Reposted by Nils Trost
aicoffeebreak.bsky.social
Just say “Wait…” – and your LLM gets smarter?!
We explain how just 1,000 training examples + a tiny trick at inference time = o1-preview level reasoning. No RL, no massive data needed.
🎥 Watch now → youtu.be/XuH2QTAC5yI
s1: Simple test-time scaling: Just “wait…” + 1,000 training examples? | PAPER EXPLAINED
YouTube video by AI Coffee Break with Letitia
youtu.be
Reposted by Nils Trost
kaessmannlab.bsky.social
How does gene regulation shape brain evolution? Our new preprint dives into this question in the context of mammalian cerebellum development! rb.gy/dbcxjz
Led by @ioansarr.bsky.social, @marisepp.bsky.social and @tyamadat.bsky.social, in collaboration with @steinaerts.bsky.social
Reposted by Nils Trost
aicoffeebreak.bsky.social
An educational and a bit historical deep dive into LLM research.
💡Learn what breakthroughs since 2017 paved the way for AI like ChatGPT (it wasn't overnight). We go through:
* Transformers
* Prompting
* Human Feedback, etc. and break it all down for you! 👇

📺 youtu.be/BprirYymXrg
LLMs Explained: A Deep Dive into Transformers, Prompts, and Human Feedback
YouTube video by AI Coffee Break with Letitia
youtu.be
Reposted by Nils Trost
aicoffeebreak.bsky.social
New video about:
REPA (Representation Alignment), a clever trick to align diffusion transformers’ representations with pretrained transformers like DINOv2.

It accelerates training and improves the diff. model’s ability to do things other than image generation (like image classification).
REPA Representation Alignment for Generation: Training Diffusion Transformers Is Easier Than You ...
YouTube video by AI Coffee Break with Letitia
youtu.be