Tessera
Folgen

Philine Lou Bommer

@philinelb.bsky.social

PhD - interpretability X climate science @TUBerlin @ATBPotsdam

10 Posts125 Follower46 Folgt
Laura Kopf@lkopf.bsky.social · 9mo ago

📄 openreview.net/forum?id=btJ... Special thanks to our co-authors @nfel.bsky.social @kirillbykov.bsky.social @philinelb.bsky.social Anna Hedström, and Marina Höhne who are not with us in person at the conference.

Capturing Polysemanticity with PRISM: A Multi-Concept Feature...

Automated interpretability research aims to identify concepts encoded in neural network features to enhance human understanding of model behavior. Within the context of large language models (LLMs)...

openreview.net

Philine Lou Bommer@philinelb.bsky.social · 11mo ago

What an amazing work and team!🎉 We are excited to discus’s Polysemanticity @neuripsconf.bsky.social with you! 🧠 Check out the paper at the link below!

Laura Kopf@lkopf.bsky.social · 11mo ago

Happy to share that our PRISM paper has been accepted at #NeurIPS2025 🎉 In this work, we introduce a multi-concept feature description framework that can identify and score polysemantic features. 📄 Paper: arxiv.org/abs/2506.15538 #NeurIPS #MechInterp #XAI

Philine Lou Bommer@philinelb.bsky.social · last yr.

🚨New paper 🚨 We are happy to announce that our paper “Deep Learning meets Teleconnections: Improving S2S Predictions for European Winter Weather” has been published at Machine Learning: Earth @ioppublishing.bsky.social 📄 iopscience.iop.org/article/10.1... 💻 github.com/philine-bomm...

Radware Bot Manager Captcha

To ensure we keep this website safe, please can you confirm you are a human by ticking the box below.

iopscience.iop.org

Philine Lou Bommer@philinelb.bsky.social · last yr.

🚨Preprint alert 🚨 Why limit our view to a single concept when neurons potentially encode multiple concepts? 👀 Check out 🔎PRISM🔎, our new multi-concept interpretability framework for LLMs. Congratulations to @lkopf.bsky.social and the team! 📄: arxiv.org/abs/2506.15538

Laura Kopf@lkopf.bsky.social · last yr.

🔍 When do neurons encode multiple concepts? We introduce PRISM, a framework for extracting multi-concept feature descriptions to better understand polysemanticity. 📄 Capturing Polysemanticity with PRISM: A Multi-Concept Feature Description Framework arxiv.org/abs/2506.15538 🧵 (1/7)

Philine Lou Bommer@philinelb.bsky.social · last yr.

🚨new preprint alert🚨 Can deep learning boost S2S forecasts? We investigated how to leverage teleconnections to improve deep learning predictions! #AI #Weather #Climate 📄: arxiv.org/abs/2504.07625 💻: (coming soon) github.com/philine-bomm...

Deep Learning Meets Teleconnections: Improving S2S Predictions for European Winter Weather

Predictions on subseasonal-to-seasonal (S2S) timescales--ranging from two weeks to two month--are crucial for early warning systems but remain challenging owing to chaos in the climate system. Telecon...

arxiv.org