Max Burg's Avatar

Max Burg

@maxfburg.bsky.social

95 Followers  |  504 Following  |  1 Posts  |  Joined: 03.02.2025
Posts Following

Posts by Max Burg (@maxfburg.bsky.social)

Post image

🚨 New paper alert! 🚨
We’ve just launched openretina, an open-source framework for collaborative retina modeling across datasets and species.
A πŸ§΅πŸ‘‡ (1/9)

14.03.2025 09:41 β€” πŸ‘ 38    πŸ” 20    πŸ’¬ 1    πŸ“Œ 1

Deadline extended to March 19 for the EVAL-FoMo workshop @cvprconference.bsky.social! We welcome submissions (incl. published papers) analyzing emerging capabilities & limits in visual foundation models.

Details: sites.google.com/view/eval-fo...
#CVPR2025

12.03.2025 12:20 β€” πŸ‘ 3    πŸ” 1    πŸ’¬ 0    πŸ“Œ 0
Post image

New preprint out! πŸŽ‰

How does LLM training loss translate to downstream performance?

We show that pretraining data and tokenizer shape loss-to-loss scaling, while architecture and other factors play a surprisingly minor role!
brendel-group.github.io/llm-line/ 🧡1/8

18.02.2025 14:09 β€” πŸ‘ 18    πŸ” 8    πŸ’¬ 1    πŸ“Œ 2

Cool paper, congrats!

18.02.2025 08:54 β€” πŸ‘ 2    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0
Video thumbnail

🧠 Meet LAMINR: Where neurons meet their matchβ€”finding common patterns that keep them excited πŸ”₯πŸ’˜

LAMINR learns & aligns invariance manifolds of sensory neurons to reveal shared properties.

🌟 #ICLR2025 Oral
πŸ“ Paper: shorturl.at/08968
πŸ’» Code: github.com/sinzlab/laminr

17.02.2025 16:51 β€” πŸ‘ 17    πŸ” 6    πŸ’¬ 2    πŸ“Œ 1

CuratedThoughts: Data Curation for RL Datasets πŸš€

Since DeepSeek-R1 introduced reasoning-based RL, datasets like Open-R1 & OpenThoughts emerged for fine-tuning & GRPO. Our deep dive found major flaws β€” 25% of OpenThoughts needed elimination by data curation.

Here's why πŸ‘‡πŸ§΅

17.02.2025 18:22 β€” πŸ‘ 13    πŸ” 9    πŸ’¬ 1    πŸ“Œ 1
Improvements in LM (top) and EM (bottom) of our micro-sam model (finetuned) compared to the default SAM model.

Improvements in LM (top) and EM (bottom) of our micro-sam model (finetuned) compared to the default SAM model.

After a long journey, Segment Anything for Microscopy is now published in Nature Methods! We significantly improve SAM for interactive and automatic segmentation in light and electron microscopy and build a user-friendly tool.
www.nature.com/articles/s41...

12.02.2025 11:41 β€” πŸ‘ 351    πŸ” 120    πŸ’¬ 11    πŸ“Œ 7
Hiring announcement: ELLIS Institute TΓΌbingen is looking for ML Researchers & Engineers for Open-Source AI Tutoring (m/f/d). The image features a white background with bold black text and the colorful ELLIS logo at the bottom.

Hiring announcement: ELLIS Institute TΓΌbingen is looking for ML Researchers & Engineers for Open-Source AI Tutoring (m/f/d). The image features a white background with bold black text and the colorful ELLIS logo at the bottom.

πŸš€ We’re hiring! Join Bernhard SchΓΆlkopf & me at @ellisinsttue.bsky.social to push the frontier of #AI in education!

We’re building cutting-edge, open-source AI tutoring models for high-quality, adaptive learning for all pupils with support from the Hector Foundation.

πŸ‘‰ forms.gle/sxvXbJhZSccr...

11.02.2025 16:34 β€” πŸ‘ 9    πŸ” 14    πŸ’¬ 1    πŸ“Œ 1