iiinova — where your digital self keeps evolving

Reposted by Sid

Computer Vision and Machine Learning at MPI Informatics @cvml.mpi-inf.mpg.de · 07/09/2026

🎉 Our group is heading to #ECCV2026 in Malmö (September 8–12) with five papers across the main conference and workshops, including one Spotlight! Come find us at the posters, and a big thank you to all our collaborators!

063

Sid @sidgairo18.bsky.social · 09/03/2026

Super, thanks! Looking very much forward to it 😃

000

Sid @sidgairo18.bsky.social · 09/03/2026

Any plans for having a 2nd iteration of the CVPR@Paris meeting 🤔? Given that last year's event was great and very much appreciated it does call for another one 😉 @davidpicard.eurosky.social ?

120

Sid @sidgairo18.bsky.social · 10/02/2026

11/11 This is joint work with Adam Wróbel (project lead), Jacek Tabor, Bernt Schiele, Bartosz Zieliński, and Dawid Rymarczyk. 📰Preprint: arxiv.org/abs/2602.06613 💻Code: github.com/a-vrobell/DAVE [to be released soon]

arxiv.org

DAVE: Distribution-aware Attribution via ViT Gradient Decomposition

Vision Transformers (ViTs) have become a dominant architecture in computer vision, yet producing stable and high-resolution attribution maps for these models remains challenging. Architectural compone...

000

Sid @sidgairo18.bsky.social · 10/02/2026

10/11 🖼️ Qualitative comparison Vs prior methods, DAVE gives sharper, more object-centric attributions with less background + fewer patch-grid artifacts.

100

Sid @sidgairo18.bsky.social · 10/02/2026

9/11 🔁 Beyond standard ViTs On inherently interpretable B-cos ViTs, DAVE yields sharper, more object-aligned maps and improves localization vs built-in B-cos explanations. B-cos ViTs produce sharp attributions only when trained with a conv-stem, DAVE fixes this reliance.

100

Sid @sidgairo18.bsky.social · 10/02/2026

8/11 📈 Main results Consistent improvements on localization metrics and perturbation evaluations.

100

Sid @sidgairo18.bsky.social · 10/02/2026

7/11 🔧 The pipeline Sample small spatial transforms + noise, compute effective transformation (conditioned forward blocks gradients through conditioning), inverse-transform & average, then apply element-wise to the input.

100

Sid @sidgairo18.bsky.social · 10/02/2026

6/11 🌫️ Stabilize (low-pass) DAVE adds low-pass stabilization by averaging the equivariant effective transformation under small input perturbations (Gaussian smoothing in expectation). This removes components unstable to tiny input changes. (See above fig., last column.)

100

Sid @sidgairo18.bsky.social · 10/02/2026

5/11 🧼 Remove grid artifacts Even the effective transformation can carry architecture-induced grid patterns. DAVE filters them by enforcing local equivariance: under small spatial transforms, the attribution must transform consistently.

100

Sid @sidgairo18.bsky.social · 10/02/2026

4/11 ⚠️ Why gradients break Operator variation can amplify tiny perturbations → high-frequency junk in attributions. DAVE drops this term and keeps the effective transformation as a cleaner attribution operator.

100

Sid @sidgairo18.bsky.social · 10/02/2026

3/11 💡 Key idea Model each ViT layer as an input-dependent linear operator L(X) applied to X. Then the input-gradient decomposes into: (1) effective transformation L(X) (2) operator variation (how L changes w.r.t. X)

100

Sid @sidgairo18.bsky.social · 10/02/2026

2/11 🧩 The problem ViT components (patch embedding, attention routing, etc.) inject structured artifacts into gradients → explanations become noisy/unstable, or methods fall back to coarse patch-level maps.

100

Sid @sidgairo18.bsky.social · 10/02/2026

🚀New preprint: DAVE — Distribution-aware Attribution via ViT Gradient DEcomposition. 1/11 🔍 What’s new: We fix a persistent issue in ViT explainability: unstable, artifact-heavy pixel attributions. DAVE yields fine-grained pixel-level maps without patch-grid saliency.

111

Reposted by Sid

Computer Vision and Machine Learning at MPI Informatics @cvml.mpi-inf.mpg.de · 19/10/2025

ICCV 2025 🌺 Aloha from Hawaii! MPI-INF (D2) is presenting 4 papers this year (one Highlight). Thread 👇

1126

Sid @sidgairo18.bsky.social · 09/10/2025

Heading to #ICCV2025? Don’t miss #CV4DC — Oct 19, Honolulu 🌺 An in-person workshop spotlighting research for/with developing regions. See exciting keynote talks from leading researches! See accepted papers & posters here: cv4dc.github.io/2025/ #AI4Good

000

Sid @sidgairo18.bsky.social · 12/08/2025

Hope these help you as much as they’ve helped me. 📚Feel free to share, bookmark, and contribute here (github.com/sidgairo18/s...)! P.S.: Please feel free to share relevant resources in the comments / thread and I'll add them as well 😀 (n/n)

github.com

GitHub - sidgairo18/sidgairo18.github.io

Contribute to sidgairo18/sidgairo18.github.io development by creating an account on GitHub.

000

Sid @sidgairo18.bsky.social · 12/08/2025

3️⃣ How to Write Academic Papers A practical checklist-driven guide on writing with clarity, rigor, and reproducibility. Inspired by ICML's best practices and more. 🔗https://sidgairo18.github.io/how_to_write_academic_papers.html (4/n)

100

Sid @sidgairo18.bsky.social · 12/08/2025

2️⃣ How to Review Scientific Papers What makes a good review? This guide compiles best practices from ICML, ICLR, CVPR, and other leading conferences. 🔗https://sidgairo18.github.io/how_to_review_scientific_papers.html (3/n)

100

Sid @sidgairo18.bsky.social · 12/08/2025

1️⃣ How to Do Research Mindset, habits, tools, writing, productivity, and advice. 🔗https://sidgairo18.github.io/notes_and_resources_on_how_to_do_research.html (2/n)

110

Sid @sidgairo18.bsky.social · 12/08/2025

🌟Here are some resources and notes about doing research that I’ve been compiling since I started my PhD.🌟 I rely on these often, and after sharing them with a few folks in my group who found them useful, I thought it could be of use to the broader community 🧵👇 (1/n)

160

Sid @sidgairo18.bsky.social · 24/07/2025

⏳Still need to wait for your last experiment results? 📣 We're pleased to announce that the deadline for non-proceeding track #CV4DC at @iccv.bsky.social has been extended to August 15, 2025 Looking forward to your submissions! cv4dc.github.io/2025/

022

Reposted by Sid

Computer Vision and Machine Learning at MPI Informatics @cvml.mpi-inf.mpg.de · 13/07/2025

Papers being presented from our group at #ICML2025! Congratulations to all the authors! To know more, visit us in the poster sessions! A 🧵with more details: @icmlconf.bsky.social @mpi-inf.mpg.de

Papers accepted at ICML 2025 from the Computer Vision and Machine Learning Department at the Max Planck Institute for Informatics.

1215

Sid @sidgairo18.bsky.social · 12/07/2025

📣 Proceeding track's results are out. 🎉 Congratulations to all the authors whose papers were accepted. We can't wait to meet you at @iccv.bsky.social in Hawaii on Oct 19th. ⏰ Our non-proceeding track is still accepting submissions until July 20th! Details in the comments

172

Sid @sidgairo18.bsky.social · 02/07/2025

Porto, Portugal 🇵🇹 June-July 2025 One of the prettiest and liveliest cities I’ve traveled to ✨ 🚤

010

Sid @sidgairo18.bsky.social · 26/06/2025

Congratulations Matt! 🔥

000

Sid @sidgairo18.bsky.social · 23/06/2025

Easy to catch on smaller datasets / experiments (eg CIFAR100). 🚨

100

Sid @sidgairo18.bsky.social · 23/06/2025

🚨More tiny experiments that seem benign: linear probing on frozen features (but again not explicitly making BatchNorm params to .eval) can have big implications. Best test accuracy: 71.20% at Epoch 10, for Learning Rate = 0.01 Makes a hell of a lot of difference! Almost 10% 🚨

100

Sid @sidgairo18.bsky.social · 23/06/2025

Seemingly innocuous error that I see students making time and again: model.eval() or param.eval() for models with norm layers (like batch-norm) is critical. Especially when freezing ❄️ certain layers and training rest.

000

Reposted by Sid

Computer Vision and Machine Learning at MPI Informatics @cvml.mpi-inf.mpg.de · 11/06/2025

🎉 Exciting News #CVPR2025! We’re proud to announce that we have 5 papers accepted to the main conference and 7 papers accepted at various CVPR workshops this year! We’re looking forward to sharing our research with the community in Nashville! Stay tuned for more details! ‪‪@mpi-inf.mpg.de‬

062

Sid @sidgairo18.bsky.social · 15/05/2025

Had my older MTB catching dust in the bike room, took 2 hours to replace both tyres and tubes with brand new ones, new brakes, and chain cleaned + greased. 💥 Pretty relaxing and a nice rejuvinating break from research✨ P.S.: Once you are used to driving road bikes, MTBs seem too slow 🤪

020

Sid @sidgairo18.bsky.social · 14/05/2025

Given the volume & scale of new work coming out, noise is going to exist. Wouldn’t these “bad actors”/“weeds” be removed organically over time when their work isn’t built on / used further? Also, how prevalent were such things before, potentially easier to pinpoint back then given a smaller scale?🤔

000

Sid @sidgairo18.bsky.social · 04/05/2025

Had a great time presenting our work at ICLR 2025 in Singapore 🇸🇬. Lot of great interactions, meeting amazing people and a very well organised conference 🌟. It was great to catch up with old colleagues and meet new folks. For more detail see: arxiv.org/abs/2503.00641

030

Sid @sidgairo18.bsky.social · 16/04/2025

🚀 I will be @iclr_conf in Singapore next week presenting our recent work. We answer how exactly does "What you explain depend on how you train". 🤓 We will be @ Hall 3 + Hall 2B #492 (26th April, 3PM - 5:30 PM) Please reach out via dm if interestd in a chat or ☕️ #ICLR2025

061

Reposted by Sid

Computer Vision and Machine Learning at MPI Informatics @cvml.mpi-inf.mpg.de · 11/04/2025

Congratulations to Max Losch for successfully defending his PhD! 🥳🎉🎓 In his thesis, "Improving Trustworthiness of Deep Learning via Inspectable and Robust Representations", he explores transparency and robustness of deep networks. We wish him the best!

072

Reposted by Sid

Dmytro Mishkin @ducha-aiki.bsky.social · 13/04/2025

Reminder that semantic segmentation evaluation numbers are dependent on the image resolution…but less than one may think (see 4x and 8x downsampling results. Source: “Fully Convolutional Networks for Semantic Segmentation” arxiv.org/abs/1411.403...

1121

Reposted by Sid

Computer Vision and Machine Learning at MPI Informatics @cvml.mpi-inf.mpg.de · 09/04/2025

Hello world, we are now on Bluesky 🦋! Follow us to receive updates on exciting research and projects from our group! #computervision #machinelearning #research

0114

Sid @sidgairo18.bsky.social · 08/04/2025

Spring beckons ☀️🪻

010

Sid @sidgairo18.bsky.social · 05/04/2025

Morning run 🏃🏻‍♂️ ☀️

010

Reposted by Sid

Serge Belongie @serge.belongie.com · 03/04/2025

So far, 2,135 people have responded to the poll Søren and I posted a few days ago. Of those, 94.4% replied “Yes” to being interested in officially presenting accepted @neuripsconf.bsky.social papers in Europe. (1/7)

57924