Associate Professor & Co-Founder - Dynamical Deep Learning
Amsterdam, The Netherlands
Joined September 2011
- Tweets954
- Following586
- Followers2.8K
- Likes2.4K
Efstratios Gavves retweeted
🛗 Introducing LiFT: A flow matching transformer that loops at inference far beyond its training depth, with no retraining, early exits, or other modifications. 💡
♾️ LiFT: Loop Flow Transformers 🧵👇
Diffusion models usually spend more inference compute by taking more sampling steps. What if we could also spend compute inside each step?
LiFT turns part of a DiT into a recurrent core, letting us scale depth at inference with only a small architectural change. On ImageNet, it beats dense DiT-XL/2 with 3.34 lower FID, while using 60% fewer parameters, 32% fewer training FLOPs, and 52% fewer inference FLOPs.
Jointly led with Mohammad Mahdi Derakhshani 👑(@mmderakhshani)
And amazing supervisors Cees Snoek (@cgmsnoek), Jan-Willem van de Meent (@jwvdm) and Gertjan J. Burghouts (@gjburghouts)
Great article, this line of work of “learnable digital twins”/“learnable simulators” is what we first pioneers last year with DreMa 😃
And there is more coming 😃
dreamtomanipulate.github.io/
From World Models To Robot World Models!! Please come join us into our next workshop held in RSS in Sydney!
#physicalai #worldmodels
🚀 Announcing #RSS2026 workshop: Robot World Models 🌐🤖🐨🌐 July 17, 2026 in 🇦🇺!
We're bringing together researchers building the next generation of WM for robots — models robots can act in: controllable, long-horizon & grounded in real dynamics. 🔗 robot-worldmodels.github.io
Efstratios Gavves retweeted
🚀 Announcing #RSS2026 workshop: Robot World Models 🌐🤖🐨🌐 July 17, 2026 in 🇦🇺!
We're bringing together researchers building the next generation of WM for robots — models robots can act in: controllable, long-horizon & grounded in real dynamics. 🔗 robot-worldmodels.github.io
Efstratios Gavves retweeted
60 years of exponential growth in chip density was achieved not through one breakthrough or technology, but a series of problems solved and new paradigms explored as old ones hit limits.
I don't think current AI has hit a wall, but even if it does, there many paths forward now.
Efstratios Gavves retweeted
Are you interested in the interdisciplinary PhD in the intersection of Machine Learning and Bioinformatics in @UvA_Amsterdam and @prinsesmaximac, working with @egavves, @IamAnneRios, and me?
More details and application: werkenbij.uva.nl/en/vacancie…
Deadline: 31 August 2025
The next wave of AI will be cyberphysical, embodied, empowered with robot worlds!
Super psyched for our first workshop on robot world models, marrying Computer Vision, Robot Learning, and Deep Learning, and with a dream team of co-organizers including Toyota Robotics.
🚀 We’re excited to announce our #CoRL2025 workshop: Learning to Simulate Robot Worlds
Spanning high-fidelity simulators, digital twins, and learned world models - our goal is to unite communities to push robot learning forward 🤖🌐
🔗 simulatingrobotworlds.github…
🧵🧵🧵
Efstratios Gavves retweeted
🚀 We’re excited to announce our #CoRL2025 workshop: Learning to Simulate Robot Worlds
Spanning high-fidelity simulators, digital twins, and learned world models - our goal is to unite communities to push robot learning forward 🤖🌐
🔗 simulatingrobotworlds.github…
🧵🧵🧵
Efstratios Gavves retweeted
Big thanks to all the organizers who are helping to make this workshop a success, and see you in 🇰🇷Seoul.
🚀🚀🚀
P.S. More detailed information about speakers and submissions is coming soon!
🔗 simulatingrobotworlds.github…
Efstratios Gavves retweeted
I have not vouched for or promoted any other event in 10 years of existing on twitter. I broke my policy for this one. It’s not the AI part that makes this conference unique. It’s the collective brain power that is off the charts.
Or later ICCV2025 on temporal self Sorenson, check it out!
New paper out - accepted at
@ICCVConference
We introduce MoSiC, a self-supervised learning framework that learns temporally consistent representations from video using motion cues.
Key idea: leverage long-range point tracks to enforce dense feature coherence across time.🧵
Efstratios Gavves retweeted
New paper out - accepted at
@ICCVConference
We introduce MoSiC, a self-supervised learning framework that learns temporally consistent representations from video using motion cues.
Key idea: leverage long-range point tracks to enforce dense feature coherence across time.🧵
Efstratios Gavves retweeted
New paper accepted at @ICCVConference: MoSiC, our new post-pretraining technique for upgrading vision foundation models like DINOv2R using videos, thanks to strong point trackers and Sinkhorn clustering.
Check the thread below :)
New paper out - accepted at
@ICCVConference
We introduce MoSiC, a self-supervised learning framework that learns temporally consistent representations from video using motion cues.
Key idea: leverage long-range point tracks to enforce dense feature coherence across time.🧵
An excellent work by @ZadaianchukML , it's not every year that an MSc project gets a CVPR oral!!
How to represent dynamic real-world data both consistently and efficiently, while reflecting the compositional object-centric structure of the world?
Contrast your slots!
...with our new SlotContrast method(🚀#CVPR2025 Oral🚀)!
🌐website: slotcontrast.github.io
🧵🧵🧵 1/n
Efstratios Gavves retweeted
How to represent dynamic real-world data both consistently and efficiently, while reflecting the compositional object-centric structure of the world?
Contrast your slots!
...with our new SlotContrast method(🚀#CVPR2025 Oral🚀)!
🌐website: slotcontrast.github.io
🧵🧵🧵 1/n
Efstratios Gavves retweeted
To all compatriots and friends of Greece: submit your most recent (already) published paper or a new idea to our inaugural Greeks in AI conference. Our first scientific get-together does not count as an archival event and it is a great opportunity to show your work to the rapidly growing AI community and ecosystem of Greece. Submit your work and register for the conference (only 20E for students) ! We have an amazing line-up of speakers, you will be able to present your submitted work with a poster, and meet new friends.
🇬🇷 Greeks in AI is gaining traction!
✅ 200+ sign-ups
📄 30+ OpenReview submissions
🤝 Growing sponsor list
🚨 Limited space @ Serafeio — register now: greeksin.ai/
Stay tuned for updates on program, speakers & more!
#GreeksInAI #AI #ML #Research #Greece #OpenReview
🇬🇷 Greeks in AI is gaining traction!
✅ 200+ sign-ups
📄 30+ OpenReview submissions
🤝 Growing sponsor list
🚨 Limited space @ Serafeio — register now: greeksin.ai/
Stay tuned for updates on program, speakers & more!
#GreeksInAI #AI #ML #Research #Greece #OpenReview
I agree, this year the due process didn't make sense. In my case, I got a paper rejected with a single WR, whose reviewer didn't even acknowledge the rebuttal (!!!); what is the point of confirmed then?
