Zhejian Yang
Papers
1
Total Citations
2
H-Index
1
About
Zhejian Yang is a rising researcher at the forefront of continual reinforcement learning and generative modeling. His work focuses on bridging the gap between static, offline training paradigms and the dynamic, ever-changing demands of real-world AI systems. Yang’s major contribution, the **Continual Diffuser (CoD)** framework, introduces a novel experience rehearsal mechanism that enables diffusion-based models to master sequential tasks without catastrophic forgetting. This breakthrough is critical for applications like robotic control, where agents must adapt to shifting environments without retraining from scratch. While his most-cited paper, *"Continual Diffuser (CoD): Mastering Continual Offline RL With Experience Rehearsal"* (2025), has already garnered early attention with 2 citations, its impact is poised to grow as the field grapples with lifelong learning challenges. Yang’s work stands out for its elegant synthesis of diffusion models and continual learning principles, offering a practical path toward more resilient, autonomous systems. His research is particularly notable for addressing a core limitation of modern AI—static training data—and pushing toward agents that learn and evolve continuously, much like their biological counterparts.
Research Focus
Key Achievements
Top Papers
- 1