Karsten Roth

Karsten Roth

@confusezius.bsky.social

Large Models, Multimodality, Continual Learning | ELLIS ML PhD with Oriol Vinyals & Zeynep Akata | Previously Google DeepMind, Meta AI, AWS, Vector, MILA 🔗 karroth.com

1.3K followers261 following16 posts

Karsten Roth
@confusezius.bsky.social

How does lifelong knowledge editing currently hold up in the real world? Fun new work probing where we are at these days with injecting new knowledge into LLMs!

Lukas Thede@lukasthede.bsky.social

🧠 Keeping LLMs factually up to date is a common motivation for knowledge editing. But what would it actually take to support this in practice at the scale and speed the real world demands? We explore this question and really push the limits of lifelong knowledge editing in the wild. 👇

Karsten Roth reposted

Zeynep Akata
@zeynepakata.bsky.social

Replying to @zeynepakata.bsky.social

📄 Disentangled Representation Learning with the Gromov-Monge Gap with Théo Uscidda, Luca Eyring, @confusezius.bsky.social, Fabian J Theis, Marco Cuturi 📄 Decoupling Angles and Strength in Low-rank Adaptation with Massimo Bini, Leander Girrbach

Karsten Roth reposted

Luca Eyring
@lucaeyring.bsky.social

Can we enhance the performance of T2I models without any fine-tuning? We show that with our ReNO, Reward-based Noise Optimization, one-step models consistently surpass the performance of all current open-source Text-to-Image models within the computational budget of 20-50 sec! #NeurIPS2024

Karsten Roth
@confusezius.bsky.social

How far can you push model merging over time, as more experts and options to model-merge arise? We comprehensively and systematically investigate this in our new work, check it out!

Sebastian Dziadzio@dziadzio.bsky.social

📄 New Paper: "How to Merge Your Multimodal Models Over Time?" arxiv.org/abs/2412.06712 Model merging assumes all finetuned models are available at once. But what if they need to be created over time? We study Temporal Model Merging through the TIME framework to find out! 🧵

arxiv.orgHow to Merge Your Multimodal Models Over Time?Model merging combines multiple expert models - finetuned from a base foundation model on diverse tasks and domains - into a single, more capable model. However, most existing model merging approaches...
Karsten Roth
@confusezius.bsky.social

😵‍💫 Continually pretraining large multimodal models to keep them up-to-date all-the-time is tough, covering everything from adapters, merging, meta-scheduling to data design and more! So I'm really happy to present our large-scale study at #NeurIPS2024! Come drop by to talk about all that and more!

Karsten Roth
@confusezius.bsky.social

We will present on Wednesday - East Exhibit Hall A-C #3703 ☺️. We've also released the entire codebase with all the methods and 60+ dataloaders that can be mixed and matched in any fashion to study continual pretraining!

Karsten Roth reposted

ELLIS
@ellis.eu

🎉 Congratulations to our newly accepted ELLIS Fellows & Scholars in 2024! Top researchers in #MachineLearning join the network to advance science & mentor the next generation. #ELLISforEurope #AI 🌍 Know someone on the list? bit.ly/3ZJd9Cz Tag them in a reply with congratulations.

bit.ly161 outstanding machine learning researchers accepted as new ELLIS Fellows & ScholarsThe ELLIS mission is to create a diverse European network that promotes research excellence and advances breakthroughs in AI, as well as a pan-European PhD program to educate the next generation of AI...

Karsten Roth reposted

Dima Damen @ECCV 2026
@dimadamen.bsky.social

Read our paper: Context-Aware Multimodal Pretraining Now on ArXiv Can you turn vision-language models into strong any-shot models? Go beyond zero-shot performance in SigLixP (x for context) Read @confusezius.bsky.social thread below… And follow Karsten … a rising star!

Karsten Roth@confusezius.bsky.social

🤔 Can you turn your vision-language model from a great zero-shot model into a great-at-any-shot generalist? Turns out you can, and here is how: arxiv.org/abs/2411.15099 Really excited to this work on multimodal pretraining for my first bluesky entry! 🧵 A short and hopefully informative thread:

Karsten Roth reposted

Alfredo Canziani
@alfcnz.bsky.social

Beautiful paper! 😍😍😍 Captions go above the tables, but otherwise aesthetically very pleasing.

Karsten Roth@confusezius.bsky.social

🤔 Can you turn your vision-language model from a great zero-shot model into a great-at-any-shot generalist? Turns out you can, and here is how: arxiv.org/abs/2411.15099 Really excited to this work on multimodal pretraining for my first bluesky entry! 🧵 A short and hopefully informative thread:

Karsten Roth reposted

Olivier Hénaff
@olivierhenaff.bsky.social

More than zero-shot generalization, few-shot *adaptation* is critical for many applications. We find simple changes to multimodal pretraining are sufficient to yield outsized gains on a wide range of few-shot tasks. Congratulations @confusezius.bsky.social on a very successful internship!

Karsten Roth@confusezius.bsky.social

🤔 Can you turn your vision-language model from a great zero-shot model into a great-at-any-shot generalist? Turns out you can, and here is how: arxiv.org/abs/2411.15099 Really excited to this work on multimodal pretraining for my first bluesky entry! 🧵 A short and hopefully informative thread:

Karsten Roth reposted

Ivana Balazevic
@ibalazevic.bsky.social

We maintain strong zero-shot transfer of CLIP / SigLIP across model size and data scale, while achieving up to 4x few-shot sample efficiency and up to +16% performance gains! Fun project with @confusezius.bsky.social, @zeynepakata.bsky.social, @dimadamen.bsky.social and @olivierhenaff.bsky.social.

Karsten Roth@confusezius.bsky.social

🤔 Can you turn your vision-language model from a great zero-shot model into a great-at-any-shot generalist? Turns out you can, and here is how: arxiv.org/abs/2411.15099 Really excited to this work on multimodal pretraining for my first bluesky entry! 🧵 A short and hopefully informative thread:

View full thread
Karsten Roth
@confusezius.bsky.social

In LIxP, we utilize a learnable temperature separation and a simple cross-attention-based formalism to augment existing contrastive vision-language training. We teach models what to expect at test-time in few-shot scenarios.

Karsten Roth
@confusezius.bsky.social

LIxP was carefully designed and tested for scalability! LIxP also maintains the strong zero-shot transfer of CLIP and SigLIP backbones across model sizes (S to L) and data (up to 15B), and allows up to 4x sample efficiency at test time, and up to +16% performance gains!

Karsten Roth
@confusezius.bsky.social

They can struggle with applications that require operating on new context, e.g. few-shot adaptation. Why? They do not explicitly train for that! We find a surrogate objective to optimize for -- context-aware language-image pretraining (LIxP)

Karsten Roth reposted

ELLIS
@ellis.eu

Hi 👋 We're glad to be here on @bsky.app and looking forward to engaging in this community. But first, learn a little more about us... #ELLISforEurope #AI #ML #CrossBorderCollab #PhD

You are up to date.