Lukas Galke Poech

lgalke

·

https://lgalke.github.io

AI & ML interests

LLM interpretability, agentic/multi-agent safety

Recent Activity

authored a paper about 2 hours ago

Multilingual GSM-Symbolic: What determines capability transfer across languages?

upvoted a paper about 6 hours ago

Multilingual GSM-Symbolic: What determines capability transfer across languages?

authored a paper 5 days ago

Selecting The Most Informative Tokens in Natural Language Autoencoders

View all activity

Organizations

Papers 27

arxiv:2610.03367

arxiv:2609.37040

arxiv:2608.19889

arxiv:2608.13517

models 1

lgalke/Qwen3.5-35B-A3B-psysafe

Image-Text-to-Text • 36B • Updated Mar 25 • 4

datasets 0

None public yet