Sid
@sidgairo18.bsky.social
🏔️📍🇩🇪
🇪🇺 PhD student at the Max Planck Institute for Informatics, and Institute of Science & Technology - Austria.
💻🏃🏻♂️🚴🏻🏋🏻🏊⛷️🎸🎹📚
Webpage: https://sidgairo18.github.io/
🇪🇺 PhD student at the Max Planck Institute for Informatics, and Institute of Science & Technology - Austria.
💻🏃🏻♂️🚴🏻🏋🏻🏊⛷️🎸🎹📚
Webpage: https://sidgairo18.github.io/
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]
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]
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...
arxiv.org
February 10, 2026 at 12:10 PM
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]
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]
10/11 🖼️ Qualitative comparison
Vs prior methods, DAVE gives sharper, more object-centric attributions with less background + fewer patch-grid artifacts.
Vs prior methods, DAVE gives sharper, more object-centric attributions with less background + fewer patch-grid artifacts.
February 10, 2026 at 12:10 PM
10/11 🖼️ Qualitative comparison
Vs prior methods, DAVE gives sharper, more object-centric attributions with less background + fewer patch-grid artifacts.
Vs prior methods, DAVE gives sharper, more object-centric attributions with less background + fewer patch-grid artifacts.
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.
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.
February 10, 2026 at 12:10 PM
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.
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.
8/11 📈 Main results
Consistent improvements on localization metrics and perturbation evaluations.
Consistent improvements on localization metrics and perturbation evaluations.
February 10, 2026 at 12:10 PM
8/11 📈 Main results
Consistent improvements on localization metrics and perturbation evaluations.
Consistent improvements on localization metrics and perturbation evaluations.
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.
Sample small spatial transforms + noise, compute effective transformation (conditioned forward blocks gradients through conditioning), inverse-transform & average, then apply element-wise to the input.
February 10, 2026 at 12:10 PM
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.
Sample small spatial transforms + noise, compute effective transformation (conditioned forward blocks gradients through conditioning), inverse-transform & average, then apply element-wise to the input.
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.)
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.)
February 10, 2026 at 12:10 PM
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.)
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.)
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.
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.
February 10, 2026 at 12:10 PM
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.
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.
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.
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.
February 10, 2026 at 12:10 PM
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.
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.
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)
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)
February 10, 2026 at 12:10 PM
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)
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)
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.
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.
February 10, 2026 at 12:10 PM
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.
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.
The Great Depression, Recession and Stagnation in Full Historical Context
www.cold-takes.com
August 13, 2025 at 9:04 PM
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)
📚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 - sidgairo18/sidgairo18.github.io
Contribute to sidgairo18/sidgairo18.github.io development by creating an account on GitHub.
github.com
August 12, 2025 at 2:10 PM
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)
📚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)
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)
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)
August 12, 2025 at 2:10 PM
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)
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)
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)
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)
August 12, 2025 at 2:10 PM
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)
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)
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)
Mindset, habits, tools, writing, productivity, and advice.
🔗https://sidgairo18.github.io/notes_and_resources_on_how_to_do_research.html (2/n)
August 12, 2025 at 2:10 PM
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)
Mindset, habits, tools, writing, productivity, and advice.
🔗https://sidgairo18.github.io/notes_and_resources_on_how_to_do_research.html (2/n)
Submit your papers to our non-proceeding track here: cv4dc.github.io/2025/call-fo...
Computer Vision for Developing Countries (CV4DC) Workshop 2025 - Call for Papers
cv4dc.github.io
July 12, 2025 at 9:11 AM
Submit your papers to our non-proceeding track here: cv4dc.github.io/2025/call-fo...