iiinova — where your digital self keeps evolving

Olaf Dünkel

@oduenkel.bsky.social

ELLIS PhD @ MPI & Oxford - Generative Models for Vision odunkel.github.io

Olaf Dünkel @oduenkel.bsky.social · 02/12/2025

Arrived in San Diego for #NeurIPS2025! Reach out if you are around too! Happy to chat! We'll present AttentionChains on Thursday afternoon! neurips.cc/virtual/2025...

neurips.cc

NeurIPS Poster Attention (as Discrete-Time Markov) ChainsNeurIPS 2025

000

Olaf Dünkel @oduenkel.bsky.social · 18/10/2025

Currently travelling to #ICCV2025 and looking forward to presenting DIY-SC and CNS-Bench there! DIY-SC: #538 at poster session 2 CNS-Bench: #1839 at poster session 5 Happy to chat during the poster sessions or at some other time if you are around!

150

Reposted by Olaf Dünkel

Thomas Wimmer @wimmerthomas.bsky.social · 16/10/2025

Super excited to introduce ✨ AnyUp: Universal Feature Upsampling 🔎 Upsample any feature - really any feature - with the same upsampler, no need for cumbersome retraining. SOTA feature upsampling results while being feature-agnostic at inference time. 🌐 wimmerth.github.io/anyup/

2285

Olaf Dünkel @oduenkel.bsky.social · 24/07/2025

You read this? You’ll likely read the linked post too. You like it? Your followers might see it too. In other words: Attention here → attention to Yotam’s post. We explore how transformer attention can be propagated—like PageRank, but for attention. Fun work with @yotamerel.bsky.social

yoterel.github.io

Attention (as Discrete-Time Markov) Chains

020

Olaf Dünkel @oduenkel.bsky.social · 26/06/2025

Are you using DINOv2 for tasks that require semantic features? DIY-SC might be the alternative! It refines DINOv2 or SD+DINOv2 features and achieves a new SOTA on the semantic correspondence dataset SPair-71k when not relying on annotated keypoints! [1/6] genintel.github.io/DIY-SC

1153