Sotabase

Career

· PhD student, EECS, UC Berkeley2023–

Publications (11)

Diffusion Autoencoders: Toward a Meaningful and Decodable Representation
Computer Vision and Pattern Recognition · 2021
538
cited
Guided Motion Diffusion for Controllable Human Motion Synthesis
IEEE International Conference on Computer Vision · 2023
207
cited
Optimizing Diffusion Noise Can Serve As Universal Motion Priors
Computer Vision and Pattern Recognition · 2023
68
cited
Improved image classification explainability with high-accuracy heatmaps
iScience · 2022
26
cited
Visual Jenga: Discovering Object Dependencies via Counterfactual Inpainting
arXiv.org · 2025
8
cited
CProp: Adaptive Learning Rate Scaling from Past Gradient Conformity
arXiv.org · 2019
3
cited
High resolution weakly supervised localization architectures for medical images
arXiv.org · 2020
3
cited
Set Prediction in the Latent Space
Neural Information Processing Systems · 2021
3
cited
Decom-Renorm-Merge: Model Merging on the Right Space Improves Multitasking
arXiv.org · 2025
1
cited
Accelerating Blockchain Scalability: New Models for Parallel Transaction Execution in the EVM
arXiv.org · 2025
Redesigning weakly supervised localization architectures for medical images
2023
Sotabase
Konpat Preechakul | Researcher Profile | Sotabase | Sotabase