iiinova — where your digital self keeps evolving

Matthew Kowal

@matthewkowal.bsky.social

PhD @ York University / Research Intern @ Ubisoft LaForge / Technical Lead @ VectorInst/ Previously @ Toyota Research Institute and @ NextAI Interpretability and Computer Vision

Posts Replies Media

Reposted by Matthew Kowal

Harry Thasarathan @hthasarathan.bsky.social · 07/02/2025

Our method reveals model-specific features too: DinoV2 (left) shows specialized geometric concepts (depth, perspective), while SigLIP (right) captures unique text-aware visual concepts. This opens new paths for understanding model differences! (7/9)

162

Reposted by Matthew Kowal

Kosta Derpanis @csprofkgd.bsky.social · 07/02/2025

Discover how our new mechanistic interpretability work uncovers universal concepts. Check it out on arXiv!

0403

Reposted by Matthew Kowal

Harry Thasarathan @hthasarathan.bsky.social · 07/02/2025

🌌🛰️🔭Wanna know which features are universal vs unique in your models and how to find them? Excited to share our preprint: "Universal Sparse Autoencoders: Interpretable Cross-Model Concept Alignment"! arxiv.org/abs/2502.03714 (1/9)

15617