Log inSign up
francesco croce
86 posts
@fra__31

francesco croce

@fra__31
Assistant Professor @AaltoUniversity & PI at ELLIS Institute Finland, Postdoc @tml_lab, PhD @uni_tue
Helsinki, Finland
fra31.github.io
Joined May 2019
393
Following
325
Followers
RepliesRepliesRepostsRepostsMediaMedia

Log in or sign up for X

See what’s happening and join the conversation

Continue with phone
or
Log in with username or email
Terms·Privacy·Cookies·Accessibility·Ads Info·© 2026 X Corp.
  • @fra__31
    francesco croce
    @fra__31
    Jan 30
    The 3rd Workshop on "Test-Time Updates (TTU): Putting Updates to the Test!" will be held at ICLR 2026! @iclrconf Deadline (Main and Tiny paper tracks): Feb. 6th 2026 (AoE) More info: ttu-iclr2026.github.io
    2
  • @fra__31
    francesco croce
    @fra__31
    Oct 28, 2025
    Happy to share that I've started as an assistant professor at @AaltoUniversity and ELLIS Institute Finland! I'll recruit students via the ELLIS PhD Program ellis.eu/research/phd-p… to work on multimodal learning, robustness, visual reasoning... feel free to reach out!
    Image
    4
  • @fra__31
    francesco croce
    @fra__31
    Jun 6, 2025
    We just released Chain-of-Frames: explicitly referencing relevant frames while reasoning improves the performance of video LLMs across benchmarks! Check it out 👇
    @SaraGhznfri
    Sara Ghazanfari
    @SaraGhznfri
    Jun 6, 2025
    🚨How to incorporate temporal grounding into the reasoning steps of video LLMs? 📃We’re excited to introduce Chain-of-Frames (CoF), a simple method to improve reasoning via explicit frame references! 🧠✨ Big thanks to my co-authors, @fra__31, N. Flammarion, P. Krishnamurthy,
    Image
  • @fra__31
    francesco croce
    @fra__31
    Jun 5, 2025
    📃 In our new paper, we introduce FuseLIP, an encoder for multimodal embedding. We use early fusion of modalities to train a single transformer on contrastive + masked (multimodal) modeling loss More details👇
    @chs20_
    Christian Schlarmann
    @chs20_
    Jun 5, 2025
    Excited to announce FuseLIP: an embedding model that encodes image+text into a single vector. We achieve this by tokenizing images into discrete tokens, merging these with the text tokens and subsequently processing them with a single transformer.
    Image
    1
  • @fra__31
    francesco croce
    @fra__31
    May 13, 2025
    Deadline in one week!
    @PUT_TTA_ICML25
    Workshop on Test-Time Adaptation (PUT) @ ICML2025
    @PUT_TTA_ICML25
    May 13, 2025
    ⏳We're EXACTLY a week away from the paper deadline (05/19)! RUN RUN RUN 🏃‍♂️🏃‍♀️ P.S. We'll also accept ICML papers as short forms (4 pages). tta-icml2025.github.io
Advertisement
Advertisement