mechcoder

Training a simple classifier on top of "frozen features" from large vision models is now common and powerful. In our CVPR 2024 paper (arxiv.org/abs/2403.10519), we show that just applying simple augmentations on such frozen features can improve few-shot classification.

@Pathenederland Dank je! Maar er zijn geen voorstellingen na dinsdag :-(

@mechcoder Hi! Wat leuk dat je enthousiast bent over de nieuwe Frankenstein film. De film is momenteel inderdaad nog steeds te zien in de bioscoop! Ik raad je aan te kijken voor locaties en data via de volgende link; bit.ly/4nd9J3N ✨ ^Tisha

@Pathenederland Gaan jullie Nog voorstellingen voor de film Frankenstein draaien?

@joapuipe So you didn't want to test your maths skills?

I want an extension that processes tweets and immediately mutes/blocks the account if the post includes a video like this. And no, I'm not paying for a blue check to get rid of crappy ads. Thank you in advance.

@Pathenederland ik Kan niet inloggen via de app. Kan jij het even controloren aub

I’m presenting “Agents’ Room: Narrative Generation through Multi-step Collaboration” at the #ICLR2025 poster session at 3pm today. come chat about multi-agent systems for complex writing tasks! link: iclr.cc/virtual/2025/p… paper: arxiv.org/abs/2410.02603

@ftm_guney Did you see furiosa as well?

I avoided the Mad Max movies for years, thinking they were just dumb action flicks. what a mistake. as lovely Nux (Nicholas Hoult) would say: what a movie! what a lovely movie!

@nuanced_nl @Pathenederland Please switch off your phone during the theater and don't disturb others with your selfishness

Tonight I'm finally going to watch Nezha 2 in the cinema in the Netherlands #NeZha2 @Pathenederland

@Pathenederland hoi!! Komt de "sorteren by tijd" feature weer terug?

@Pathenederland "films bekeken lijst" van de laatste 3 Jaar is kwijt. komt het terug?

@Pathenederland hoi! Waarom staat er nu geen "laatste Kans" bij de films? Ik heb bijna een para films gemist! 😄

🔥Excited to introduce RINS - a technique that boosts model performance by recursively applying early layers during inference without increasing model size or training compute flops! Not only does it significantly improve LMs, but also multimodal systems like SigLIP. (1/N)

@Pathenederland vorig jaar heb ik 200 films bekeken.

@geekmarcus @eduardo_nlp Congrats! Well done :-)

First, first-author manuscript accepted for #EMNLP2024. 🥳🥳 Big thanks to Prof. @eduardo_nlp.

Want your VLM to reflect the world's rich diversity 🌍? We’re very excited to share our recent research on this topic. TLDR: to build truly inclusive models that work for everyone, don’t filter by English, and check out our recommended evaluation benchmarks. (1/7)

@nagaraj_arvind Thank you! :-)

@mechcoder Congrats! this is such a cool idea!

Training a simple classifier on top of "frozen features" from large vision models is now common and powerful. In our CVPR 2024 paper (arxiv.org/abs/2403.10519), we show that just applying simple augmentations on such frozen features can improve few-shot classification.

Website at: frozen-feature-augmentation.github.io

This is joint work with Andreas Baer, @m__dehghani and @neilhoulsby. Check out paper arxiv.org/abs/2403.10519 for all the details!