Filip Szatkowski (@f_szatkowski) / X

Filip Szatkowski

82 posts

Filip Szatkowski

@f_szatkowski

AI Scientist at Mistral.AI and PhD Candidate at WUT

Warsaw, Poland

Joined February 2020

Filip Szatkowski
@f_szatkowski
Feb 23
We are presenting "Universal Properties of Activation Sparsity in Modern Large Language Models" at ICLR 2026! We ask a simple question: how sparse are modern LLMs, really — and does it matter? 👇
Filip Szatkowski
@f_szatkowski
Dec 13, 2024
We are presenting "Exploiting Activation Sparsity with Dense to Dynamic-k Mixture-of-Experts Conversion" together with Bartosz Wójcik and @AjPiorczynski today at #NeurIPS2024 ! 🎉 If you are around and want to chat about efficiency, stop by poster 2304 at session 6 east at 4:30!
Filip Szatkowski
@f_szatkowski
Dec 2, 2024
Heading to Vancouver 🇨🇦 next week for #NeurIPS2024! Together with @WojcikBar & @AjPiorczynski , we’ll present our work at Poster Session 6 East, #2304 (arxiv: arxiv.org/abs/2310.04361) If you are there and you are into efficiency in ML, let's connect over coffee or beer! ☕️🍺
Filip Szatkowski
@f_szatkowski
Oct 3, 2024
Good news from last week, we just got accepted to #NeurIPS2024!
Filip Szatkowski
@f_szatkowski
Jun 26, 2024
“Exploiting Activation Sparsity with Dense to Dynamic-k Mixture-of-Experts Conversion” was accepted to the ES-FoMo Workshop at #ICML2024! See the thread for more details.🧵👇 Paper co-authored with @WojcikBar @AjPiorczynski and @s_scardapane. 📄arxiv.org/abs/2310.04361
Filip Szatkowski
@f_szatkowski
Jun 26, 2024
“Exploiting Activation Sparsity with Dense to Dynamic-k Mixture-of-Experts Conversion” was accepted to the ES-FoMo Workshop at #ICML2024! See the thread for more details.🧵👇 Paper co-authored with @WojcikBar @AjPiorczynski and @s_scardapane. 📄arxiv.org/abs/2310.04361

New to X?