Posts by Collection

Continual Learning in Sequential Tasks

Exploring methods to mitigate catastrophic forgetting in sequential learning scenarios. Target: ICLR 2026

Federated Learning: Theory and Practice

Investigating convergence theory in federated learning through empirical validation. Advisor: Prof. Avi Mohan

Motion Capture and 3D Character Animation

Building end-to-end pipeline from 2D video to 3D animated characters. Advisor: Prof. Reza Rawassizadeh

3D Scene Understanding and Prediction

Developing novel approaches for world modeling combining scene reconstruction with temporal dynamics. Target: CVPR 2026

GradES: Significantly Faster Training in Transformers with Gradient-Based Early Stopping

Published in arXiv preprint, 2025

A novel gradient-based early stopping method that operates at the component level within transformer architectures, achieving 1.57x to 7.22x training speedup while improving model generalization.

Recommended citation: Qifu Wen, Xi Zeng, Zihan Zhou, Shuaijun Liu, Mehdi Hosseinzadeh, Ningxin Su, Reza Rawassizadeh. "GradES: Significantly Faster Training in Transformers with Gradient-Based Early Stopping." arXiv preprint arXiv:2509.01842, 2025.
Download Paper