Jason Rich Darmawan
AI Safety Researcher | Mechanistic Interpretability and Knowledge Editing Method for Large Language Models
Graduate Student Researcher
西北工业大学计算机学院多媒体信息处理团队（MIPLab）
Email: jasonrichdarmawan [at] gmail [dot] com
Research Interests:
- Mechanistic Interpretability and Knowledge Editing Method for Large Language Models
Selected Publications*:
- Two knowledge editing methods for large language models (LLMs) manuscripts under review.
- One mechanistic interpretability method for LLMs manuscript under review.
*Under double anonymized peer review. Email for the paper copy or schedule a meeting to get a lecture about the results.
Work Projects:
Course Projects*:
- Classifying Pediatric Brain Tumor Subtypes with Sparse Mixture of Experts Architecture
- Challenges in Federated Learning for Medical Imaging, Existing Methods, and Their Limitations
- Challenges and Properties in Deep Learning for Medical Imaging, Types of Noisy Labels, Handling Methods, and Their Limitations
*Course projects are not published in any journal or conference
Photos taken during a break at the Chang'an campus.
You can schedule meeting here and you can see the calendar for those meetings separately.
Blog post: