Sign in

Yoel Zimmermann

@yzimmermann.bsky.social
98 followers 234 following 9 posts

Exploring how the smallest shapes the biggest @ ETH Zurich. Somewhere between Physics, Chemistry, and ML.

PostsRepliesMedia
Yoel Zimmermann @yzimmermann.bsky.social · 06/01/2025
www.nature.com/articles/s41...
nature.com
Scalable watermarking for identifying large language model outputs - Nature
A scheme for watermarking the text generated by large language models shows high text quality preservation and detection accuracy and low latency, and is feasible in large-scale-production settings.
010
Reposted by Yoel Zimmermann
Alán Aspuru-Guzik @aspuru.bsky.social · 14/12/2024
What there should be is a *single* place where one deposits a standard letter for all students. This would be much more scalable and sustainable.
3163
Reposted by Yoel Zimmermann
Matt Hartings @matthartings.bsky.social · 14/12/2024
YESSS!!!
0102
Yoel Zimmermann @yzimmermann.bsky.social · 11/12/2024
P.S. If you work in the flavor/food/fragrance industry and are laughing about our 15k dataset, reach out to hello@kvantlabs.ai. We would love to build an even better model for you!
010
Yoel Zimmermann @yzimmermann.bsky.social · 11/12/2024
Preprint: chemrxiv.org/engage/chemr... Dataset: huggingface.co/datasets/Far... It has been fun working on this with @leif7ieben.bsky.social, @heseng.bsky.social, Philipp Pestlin, Franz Görlich!
chemrxiv.org
A Chemical Language Model for Molecular Taste Prediction
Determining molecular taste remains a significant challenge in food science. Here, we present FART (Flavor Analysis and Recognition Transformer), a chemical language model capable of predicting molecu...
121
Yoel Zimmermann @yzimmermann.bsky.social · 11/12/2024
Trained on the largest publicly available dataset (15,025 unique compounds), we achieve >90% accuracy - outperforming existing binary classifiers that focus on only one taste class (e.g. sweet/non-sweet). Try it out here: fartlabs-fart.hf.space
100
Yoel Zimmermann @yzimmermann.bsky.social · 11/12/2024
Ever had the sudden urge to taste something in the lab? Now you can! (For legal reasons: you can't) FART (Flavor Analysis and Recognition Transformer) is the first machine learning model to simultaneously predict four taste classes – sweet, sour, bitter, umami - from chemical structure alone!
2100
Reposted by Yoel Zimmermann
Volker Weber @vowe.net · 07/12/2024
What a time to be alive
413065378
Reposted by Yoel Zimmermann
Nathaniel Johnston @njohnston.ca · 03/12/2024
Me, a moron: I don't need a local TeX installation, Overleaf works so well and is so easy! Overleaf: like every online service that exists, goes down from time to time. Me: shocked Pikachu.
A screenshot of overleaf.com giving a "Error: Server Error" message.
001
Yoel Zimmermann @yzimmermann.bsky.social · 28/11/2024
Is 15 minutes supposed to be a lot?
110
Yoel Zimmermann @yzimmermann.bsky.social · 26/11/2024
CERN also came to my mind 😅
110
Reposted by Yoel Zimmermann
Ben Blaiszik @benblaiszik.bsky.social · 26/11/2024
For those of you looking for a big helping of LLMs for Thanksgiving, here you go! 🤖🦃 📝 arxiv.org/abs/2411.15221 We release 34 examples demonstrating applications of LLMs in materials science and chemistry across: 🔶molecular and material property prediction; 🔷molecular and material design; ...
15415
Yoel Zimmermann @yzimmermann.bsky.social · 24/11/2024
In principle you can do a static evaluation on a holdout set but I have no clue if 4k datapoints is enough.
010
Yoel Zimmermann @yzimmermann.bsky.social · 24/11/2024
Cool work! How about doing RLHF on your data?
110
Reposted by Yoel Zimmermann
Andrew D Thaler @drandrewthaler.bsky.social · 04/11/2024
If only scholarly publications came with a short synopsis of the paper right up front, written by the authors to highlight the important and salient points of the study. We could give it a nifty name, like "abstract".
292168667