Who's Who of AI

Christian Wolf

341 trust researcher @chriswolfvision.bsky.social · 6,388 followers
AI research Vision & synthetic media Agents & robotics Compute & infrastructure
Why they matter

Researcher with public evidence across AI research, Vision & synthetic media, Agents & robotics.

AI signals
5
past 30d
Sources
5
distinct domains
Discussions
36
past 30d
Latest signal
4d ago
View every signal from Christian Wolf →
Principal Scientist at Naver Labs Europe, Lead of Spatial AI team. AI for Robotics, Computer Vision, Machine Learning. Austrian in France. https://chriswolfvision.github.io/www/
What they're sharing

Articles & links

↻ Christian Wolf reposted
@pdejorge.bsky.social

1/6 Excited to share that our paper on model merging was accepted at ECCV 2026! 🎉 We introduce an efficient, decoder-free proxy that makes model selection faster, simpler and practical across vision tasks. 📄 arxiv.org/abs/2604.12935 🌐 europe.naverlabs.com/task-alignment 🧵👇

Task Alignment: A Simple Proxy for Practical Model Merging Across Diverse Vision Tasks arxiv.org
AI Weekly's analysis →
  • A NAVER Labs Europe paper accepted at ECCV 2026 introduces a 'task alignment proxy' for selecting model-merging hyperparameters without training the decoder for each candidate.
  • The method targets heterogeneous vision tasks with trainable decoders, moving beyond the CLIP image classification setup that dominated prior model merging work.
  • The authors claim the proxy speeds up hyperparameter selection by orders of magnitude while retaining downstream performance.
Read full analysis →
 View on Bluesky →
↻ Christian Wolf reposted
Dima Damen @dimadamen.bsky.social

*NEW* Our #ECCV2026 @eccv.bsky.social paper Towards in-the-wild Egocentric 3D Hand-Object Pose Estimation Now on ArXiv w Dataset, Code&model sid2697.github.io/epic-contact/ arxiv.org/abs/2606.30598 Two contributions: 1. EPIC-Contact Dataset 2. HOPformer Method &Checkpoint 🧵 1/6

Towards in-the-wild Egocentric 3D Hand-Object Pose Estimation arxiv.org
AI Weekly's analysis →
  • EPIC-Contact provides 2.3K clips and 62.3K frames of in-the-wild egocentric footage with dense, bijective 3D hand-object contact correspondences and posed meshes.
  • HOPformer is an end-to-end transformer that jointly predicts bi-manual hand and object pose in a single forward pass using a cross-attention decoder.
  • The model reaches 82.4% success rate on ARCTIC, 6.2 points above prior state of the art, and nearly doubles success rate on EPIC-Contact while cutting contact deviation by 75%.
Read full analysis →
 View on Bluesky →
↻ Christian Wolf reposted
Dmytro Mishkin @ducha-aiki.bsky.social

VGGT-Ω @jianyuanwang.bsky.social et 9 al. tl;dr: 1) simplicity to scale compute(no pcl head, no matching head, but losses kept, no rays for intrinsics, less DPT) 2) Complex video SfM/filtering procedure to scale data. 3) works great on IMC-2025(not in paper, I tested) arxiv.or…

VGGT-$Ω$ arxiv.org View on Bluesky →

Memory Caching: RNNs with growing memory Behrouz et al. arxiv.org/abs/2602.24281 Take different RNNs and, and at time t don't read out only M_t, but cached previous hidden memories. Neat and effective. Tested on different modern RNN variants.

Memory Caching: RNNs with Growing Memory arxiv.org
View on Bluesky · ♥ 28 ↻ 3 ↩ 0 · 54d ago

ECCV 2026 Workshop on Embodied Multimodal Reasoning in Physical Environments emr-workshop.github.io Sept 8th, 2026 Full paper DL has passed, but extended abstracts can still be submitted: DL July 24, 2026 (23:59 AoE)

EMR Workshop @ ECCV 2026 emr-workshop.github.io
View on Bluesky · ♥ 10 ↻ 1 ↩ 0 · 2 from the directory shared this · 25d ago
Their own posts

Recent commentary

AI bubble vs. FIFA world cup vs. Summer Heat wave

View on Bluesky · ♥ 34 ↻ 4 ↩ 11 · 31d ago

For your Embodied AI task you want a recurrent model with constant complexity per step, but you don't want to lose the power of transformers (which store the full obs history and attend to it)? Do not despair, we have your back. We distill transformers into recurrent transformers 1/8

View on Bluesky · ♥ 33 ↻ 10 ↩ 2 · 52d ago

My Dad made fun of me for my use of "please" when talking to Claude. His point: the AI does not care. My point: this is not about the AI but about me.

View on Bluesky · ♥ 27 ↻ 0 ↩ 2 · 61d ago

Excellent Keynote by @akorba.bsky.social at Cap-Rfiap in Montpellier, France, on machine learning and distances between distributions.

View on Bluesky · ♥ 17 ↻ 4 ↩ 0 · 38d ago

Goodhart's law states that "When a measure becomes a target, it ceases to be a good measure". Conference peer-review is a measure, arguably now mostly implemented through LLMs, and will become a target. Ergo, future academic research will target what Claude finds interesting and acceptable.

View on Bluesky · ♥ 17 ↻ 1 ↩ 2 · 20d ago

It is 2030 and we ran out of English phrases which we can still use because of they are not frequently employed by LLMs.

View on Bluesky · ♥ 15 ↻ 0 ↩ 3 · 25d ago

ICRA panel on the impact of AI and the paper avalanche on robotics conferences and journals. The same problems arrise in ML and in CV of course.

View on Bluesky · ♥ 15 ↻ 2 ↩ 1 · 72d ago

In 2026 the accepted papers are the ones which were reviewed by the same LLM which the authors had used to check the paper before submission.

View on Bluesky · ♥ 14 ↻ 1 ↩ 1 · 28d ago

In the book "The Diamond Age" (1995) by Neal Stephenson, a working class child jumps social classes with the help of a unique interactive book, a personal trainer which answer all her questions and allows her to learn first. For centuries we waited for this. Anti-AI crowd: AI is anti teaching!! 😡

View on Bluesky · ♥ 7 ↻ 1 ↩ 1 · 24d ago

People: trees are connected to each other and they can react to insects by releasing chemical signals. They span a big consciousness. Also people: an artificial intelligence can never ever be conscious.

View on Bluesky · ♥ 7 ↻ 1 ↩ 1 · 26d ago
Their network

In Christian Wolf's orbit

Center = Christian Wolf. Left = members they follow (green edges). Right = members who follow them (blue edges). Top = mutual follows (orange edges, slightly larger). Drag any node to reposition; click to open that profile.