Éloi Zablocki
@eloizablocki.bsky.social
Research scientist @valeoai.bsky.social
Alumnus @sorbonne-universite.fr | ENS Paris Saclay (MVA) | Ecole Polytechnique (X2012)
eloiz.github.io
Alumnus @sorbonne-universite.fr | ENS Paris Saclay (MVA) | Ecole Polytechnique (X2012)
eloiz.github.io
Reposted by Éloi Zablocki
1/🧵 Q: Can we have both a simple and SOTA architecture in autonomous driving?
R: Yes! 😍
Introducing Driving on Registers (DrivoR):
a pure Transformer backbone that achieves SOTA results in NAVSIM v1 / v2 and closed-loop HUGSIM evaluation.
Here is how 👇
R: Yes! 😍
Introducing Driving on Registers (DrivoR):
a pure Transformer backbone that achieves SOTA results in NAVSIM v1 / v2 and closed-loop HUGSIM evaluation.
Here is how 👇
January 9, 2026 at 4:55 PM
1/🧵 Q: Can we have both a simple and SOTA architecture in autonomous driving?
R: Yes! 😍
Introducing Driving on Registers (DrivoR):
a pure Transformer backbone that achieves SOTA results in NAVSIM v1 / v2 and closed-loop HUGSIM evaluation.
Here is how 👇
R: Yes! 😍
Introducing Driving on Registers (DrivoR):
a pure Transformer backbone that achieves SOTA results in NAVSIM v1 / v2 and closed-loop HUGSIM evaluation.
Here is how 👇
Reposted by Éloi Zablocki
Check out our works at @NeurIPSConf #NeurIPS2025 this week!
We present 5 full papers + 1 workshop about:
💡 self-supervised & representation learning
🖼️ generative image models
🧠 finetuning and understanding LLMs & multimodal LLMs
🔎 feature upsampling
valeoai.github.io/posts/neurip...
We present 5 full papers + 1 workshop about:
💡 self-supervised & representation learning
🖼️ generative image models
🧠 finetuning and understanding LLMs & multimodal LLMs
🔎 feature upsampling
valeoai.github.io/posts/neurip...
December 3, 2025 at 10:50 PM
Check out our works at @NeurIPSConf #NeurIPS2025 this week!
We present 5 full papers + 1 workshop about:
💡 self-supervised & representation learning
🖼️ generative image models
🧠 finetuning and understanding LLMs & multimodal LLMs
🔎 feature upsampling
valeoai.github.io/posts/neurip...
We present 5 full papers + 1 workshop about:
💡 self-supervised & representation learning
🖼️ generative image models
🧠 finetuning and understanding LLMs & multimodal LLMs
🔎 feature upsampling
valeoai.github.io/posts/neurip...
Reposted by Éloi Zablocki
1/Serve your PEFT with a fresh IPA!🍺
Finetuning large models is cheaper thanks to LoRA, but is its random init optimal?🤔
Meet IPA: a feature-aware alternative to random projections
#NeurIPS2025 WS #CCFM Oral+Best Paper
Work w/
S. Venkataramanan @tuanhungvu.bsky.social @abursuc.bsky.social M. Cord
🧵
Finetuning large models is cheaper thanks to LoRA, but is its random init optimal?🤔
Meet IPA: a feature-aware alternative to random projections
#NeurIPS2025 WS #CCFM Oral+Best Paper
Work w/
S. Venkataramanan @tuanhungvu.bsky.social @abursuc.bsky.social M. Cord
🧵
December 2, 2025 at 11:11 AM
1/Serve your PEFT with a fresh IPA!🍺
Finetuning large models is cheaper thanks to LoRA, but is its random init optimal?🤔
Meet IPA: a feature-aware alternative to random projections
#NeurIPS2025 WS #CCFM Oral+Best Paper
Work w/
S. Venkataramanan @tuanhungvu.bsky.social @abursuc.bsky.social M. Cord
🧵
Finetuning large models is cheaper thanks to LoRA, but is its random init optimal?🤔
Meet IPA: a feature-aware alternative to random projections
#NeurIPS2025 WS #CCFM Oral+Best Paper
Work w/
S. Venkataramanan @tuanhungvu.bsky.social @abursuc.bsky.social M. Cord
🧵
Off-the-shelf feature upscaling from any vision model without retraining, with NAF:
🥇 SoTA results
⚡️ High throughput
🧠 Zero-shot generalization
😍 Code available
--> NAF brings *pixel-level* features to your frozen vision backbones 🔎
👇 Check out the details!
🥇 SoTA results
⚡️ High throughput
🧠 Zero-shot generalization
😍 Code available
--> NAF brings *pixel-level* features to your frozen vision backbones 🔎
👇 Check out the details!
Need pixel-level features from your backbone (DINOv3, CLIP, RADIO, FRANCA...)?
🚀Introducing NAF: A universal, zero-shot feature upsampler.
It turns low-res ViT features into pixel-perfect maps.
-⚡ Model-agnostic
-🥇 SoTA results
-🚀 4× faster than SoTA
-📈 Scales up to 2K res
🚀Introducing NAF: A universal, zero-shot feature upsampler.
It turns low-res ViT features into pixel-perfect maps.
-⚡ Model-agnostic
-🥇 SoTA results
-🚀 4× faster than SoTA
-📈 Scales up to 2K res
November 25, 2025 at 10:55 AM
Off-the-shelf feature upscaling from any vision model without retraining, with NAF:
🥇 SoTA results
⚡️ High throughput
🧠 Zero-shot generalization
😍 Code available
--> NAF brings *pixel-level* features to your frozen vision backbones 🔎
👇 Check out the details!
🥇 SoTA results
⚡️ High throughput
🧠 Zero-shot generalization
😍 Code available
--> NAF brings *pixel-level* features to your frozen vision backbones 🔎
👇 Check out the details!
Reposted by Éloi Zablocki
CoRL 2025 is just around the corner in Seoul, Korea!
🤖 🚗
We're excited to present our latest research and connect with the community.
#CoRL2025
🤖 🚗
We're excited to present our latest research and connect with the community.
#CoRL2025
September 24, 2025 at 4:47 PM
CoRL 2025 is just around the corner in Seoul, Korea!
🤖 🚗
We're excited to present our latest research and connect with the community.
#CoRL2025
🤖 🚗
We're excited to present our latest research and connect with the community.
#CoRL2025
Reposted by Éloi Zablocki
LLM-wrapper: Black-Box Semantic-Aware Adaptation of VLMs for Referring Expression Comprehension (REC), by A. Cardiel, @eloizablocki.bsky.social, E. Ramzi, O. Simeoni, @matthieucord.bsky.social
Boost VLM perf by tuning an LLM to reason on its outputs! It's black-box🔒 & efficient⚡ (< 7h on 1 GPU)
Boost VLM perf by tuning an LLM to reason on its outputs! It's black-box🔒 & efficient⚡ (< 7h on 1 GPU)
April 9, 2025 at 9:38 AM
LLM-wrapper: Black-Box Semantic-Aware Adaptation of VLMs for Referring Expression Comprehension (REC), by A. Cardiel, @eloizablocki.bsky.social, E. Ramzi, O. Simeoni, @matthieucord.bsky.social
Boost VLM perf by tuning an LLM to reason on its outputs! It's black-box🔒 & efficient⚡ (< 7h on 1 GPU)
Boost VLM perf by tuning an LLM to reason on its outputs! It's black-box🔒 & efficient⚡ (< 7h on 1 GPU)
Reposted by Éloi Zablocki
🔥🔥🔥 CV Folks, I have some news! We're organizing a 1-day meeting in center Paris on June 6th before CVPR called CVPR@Paris (similar as NeurIPS@Paris) 🥐🍾🥖🍷
Registration is open (it's free) with priority given to authors of accepted papers: cvprinparis.github.io/CVPR2025InPa...
Big 🧵👇 with details!
Registration is open (it's free) with priority given to authors of accepted papers: cvprinparis.github.io/CVPR2025InPa...
Big 🧵👇 with details!
March 21, 2025 at 6:43 AM
🔥🔥🔥 CV Folks, I have some news! We're organizing a 1-day meeting in center Paris on June 6th before CVPR called CVPR@Paris (similar as NeurIPS@Paris) 🥐🍾🥖🍷
Registration is open (it's free) with priority given to authors of accepted papers: cvprinparis.github.io/CVPR2025InPa...
Big 🧵👇 with details!
Registration is open (it's free) with priority given to authors of accepted papers: cvprinparis.github.io/CVPR2025InPa...
Big 🧵👇 with details!
Reposted by Éloi Zablocki
🚗 Ever wondered if an AI model could learn to drive just by watching YouTube? 🎥👀
We trained a 1.2B parameter model on 1,800+ hours of raw driving videos.
No labels. No maps. Just pure observation.
And it works! 🤯
🧵👇 [1/10]
We trained a 1.2B parameter model on 1,800+ hours of raw driving videos.
No labels. No maps. Just pure observation.
And it works! 🤯
🧵👇 [1/10]
February 24, 2025 at 12:53 PM
🚗 Ever wondered if an AI model could learn to drive just by watching YouTube? 🎥👀
We trained a 1.2B parameter model on 1,800+ hours of raw driving videos.
No labels. No maps. Just pure observation.
And it works! 🤯
🧵👇 [1/10]
We trained a 1.2B parameter model on 1,800+ hours of raw driving videos.
No labels. No maps. Just pure observation.
And it works! 🤯
🧵👇 [1/10]
Reposted by Éloi Zablocki
GPUDrive got accepted to ICLR 2025!
With that, we release GPUDrive v0.4.0! 🚨 You can now install the repo and run your first fast PPO experiment in under 10 minutes.
I’m honestly so excited about the new opportunities and research the sim makes possible. 🚀 1/2
With that, we release GPUDrive v0.4.0! 🚨 You can now install the repo and run your first fast PPO experiment in under 10 minutes.
I’m honestly so excited about the new opportunities and research the sim makes possible. 🚀 1/2
February 20, 2025 at 6:53 PM
GPUDrive got accepted to ICLR 2025!
With that, we release GPUDrive v0.4.0! 🚨 You can now install the repo and run your first fast PPO experiment in under 10 minutes.
I’m honestly so excited about the new opportunities and research the sim makes possible. 🚀 1/2
With that, we release GPUDrive v0.4.0! 🚨 You can now install the repo and run your first fast PPO experiment in under 10 minutes.
I’m honestly so excited about the new opportunities and research the sim makes possible. 🚀 1/2
Reposted by Éloi Zablocki
After 6+ months in the making and over a year of GPU compute, we're excited to release the "Ultra-Scale Playbook": hf.co/spaces/nanot...
A book to learn all about 5D parallelism, ZeRO, CUDA kernels, how/why overlap compute & coms with theory, motivation, interactive plots and 4000+ experiments!
A book to learn all about 5D parallelism, ZeRO, CUDA kernels, how/why overlap compute & coms with theory, motivation, interactive plots and 4000+ experiments!
The Ultra-Scale Playbook - a Hugging Face Space by nanotron
The ultimate guide to training LLM on large GPU Clusters
hf.co
February 19, 2025 at 6:10 PM
After 6+ months in the making and over a year of GPU compute, we're excited to release the "Ultra-Scale Playbook": hf.co/spaces/nanot...
A book to learn all about 5D parallelism, ZeRO, CUDA kernels, how/why overlap compute & coms with theory, motivation, interactive plots and 4000+ experiments!
A book to learn all about 5D parallelism, ZeRO, CUDA kernels, how/why overlap compute & coms with theory, motivation, interactive plots and 4000+ experiments!
Reposted by Éloi Zablocki
We @imagineenpc.bsky.social are slowly but surely entering our proposals for master's degree internships here: docs.google.com/document/d/1...
These are 6 months projects that typically correspond to the end-of-study project in the French curriculum.
Probably more offers to come, check it regularly.
These are 6 months projects that typically correspond to the end-of-study project in the French curriculum.
Probably more offers to come, check it regularly.
2025 IMAGINE Internships
2025 Internship proposals at IMAGINE IMAGINE is a top research group on computer vision and machine learning. It is part of the LIGM lab and hosted at École des Ponts ParisTech (ENPC), about 25 min f...
docs.google.com
December 12, 2024 at 10:08 AM
We @imagineenpc.bsky.social are slowly but surely entering our proposals for master's degree internships here: docs.google.com/document/d/1...
These are 6 months projects that typically correspond to the end-of-study project in the French curriculum.
Probably more offers to come, check it regularly.
These are 6 months projects that typically correspond to the end-of-study project in the French curriculum.
Probably more offers to come, check it regularly.
🥐 Building a Computer Vision FR Starter Pack!
👉 Who else should be included?
Comment below or DM me to be added
go.bsky.app/dfvcLZ
👉 Who else should be included?
Comment below or DM me to be added
go.bsky.app/dfvcLZ
November 29, 2024 at 9:19 AM
🥐 Building a Computer Vision FR Starter Pack!
👉 Who else should be included?
Comment below or DM me to be added
go.bsky.app/dfvcLZ
👉 Who else should be included?
Comment below or DM me to be added
go.bsky.app/dfvcLZ
Reposted by Éloi Zablocki
Welcome to all the newcomers! Here's a starter pack for the 3D Computer Vision community: go.bsky.app/Cfm9XFe
November 28, 2024 at 9:30 AM
Welcome to all the newcomers! Here's a starter pack for the 3D Computer Vision community: go.bsky.app/Cfm9XFe
@csprofkgd.bsky.social Hello, thanks for maintaining the computer vision starter packs. Is it possible to be included? Thanks 🙏
November 28, 2024 at 9:05 AM
@csprofkgd.bsky.social Hello, thanks for maintaining the computer vision starter packs. Is it possible to be included? Thanks 🙏