Sotabase

Career

· Research Scientist, Reve2026–
· Member of Perception Team, Google Research
· PhD Student, UC Berkeley

Publications (33)

Deep ViT Features as Dense Visual Descriptors
arXiv.org · 2021
363
cited
“Double-DIP”: Unsupervised Image Decomposition via Coupled Deep-Image-Priors
Computer Vision and Pattern Recognition · 2018
346
cited
Visual Prompting via Image Inpainting
Neural Information Processing Systems · 2022
272
cited
Test-Time Training with Masked Autoencoders
Neural Information Processing Systems · 2022
227
cited
Explaining in Style: Training a GAN to explain a classifier in StyleSpace
IEEE International Conference on Computer Vision · 2021
174
cited
Interpreting CLIP's Image Representation via Text-Based Decomposition
International Conference on Learning Representations · 2023
151
cited
Interpreting and Editing Vision-Language Representations to Mitigate Hallucinations
International Conference on Learning Representations · 2024
66
cited
Semantic Pyramid for Image Generation
Computer Vision and Pattern Recognition · 2020
58
cited
Rosetta Neurons: Mining the Common Units in a Model Zoo
IEEE International Conference on Computer Vision · 2023
39
cited
MyStyle
ACM Transactions on Graphics · 2022
34
cited
Interpreting the Second-Order Effects of Neurons in CLIP
International Conference on Learning Representations · 2024
32
cited
Test-Time Training on Video Streams
arXiv.org · 2023
32
cited
Deep Saliency Prior for Reducing Visual Distraction
Computer Vision and Pattern Recognition · 2021
21
cited
Idempotent Generative Network
International Conference on Learning Representations · 2023
21
cited
Interpreting the Weight Space of Customized Diffusion Models
Neural Information Processing Systems · 2024
20
cited
An Empirical Study of Autoregressive Pre-training from Videos
arXiv.org · 2025
16
cited
On the Effectiveness of ViT Features as Local Semantic Descriptors
ECCV Workshops · 2022
15
cited
Vision Transformers Don't Need Trained Registers
arXiv.org · 2025
14
cited
Steering CLIP's vision transformer with sparse autoencoders
arXiv.org · 2025
13
cited
Interpreting the Repeated Token Phenomenon in Large Language Models
International Conference on Machine Learning · 2025
9
cited
Sotabase
Yossi Gandelsman | Researcher Profile | Sotabase | Sotabase