News & Milestones
Paper Accepted — Findings of EMNLP 2026
Our paper, “Surgical Alignment in Knowledge Graph Training for Clinical Diagnosis with Large Language Models,” by Saksham Khatwani, He Cheng, Majid Afshar, Dmitriy Dligach, and Yanjun Gao, has been accepted to Findings of EMNLP 2026. Paper · Code
Paper Accepted — ACL 2026
Our paper, “LogosKG: Hardware-Optimized Scalable and Interpretable Knowledge Graph Retrieval,” was accepted to the 64th Annual Meeting of the Association for Computational Linguistics. The paper, code, and online demo are publicly available.
BadSAD Published — IEEE Big Data 2025
Our paper, “BadSAD: Clean-Label Backdoor Attacks against Deep Semi-Supervised Anomaly Detection,” appeared in the proceedings of the 2025 IEEE International Conference on Big Data. Paper
Joined CU Anschutz as a Postdoctoral Researcher
Joined the Department of Biomedical Informatics at the University of Colorado Anschutz Medical Campus. My work focuses on biomedical knowledge-graph retrieval, large language model reasoning, and multi-agent systems for clinical applications.
Ph.D. Dissertation Defense — Completed
Successfully defended my dissertation, “Interpretable and Robust Deep Anomaly Detection,” and completed the Ph.D. in Computer Science at Utah State University.
Volunteer Lead — IEEE Big Data 2024
Served as Volunteer Lead for the 2024 IEEE International Conference on Big Data.
Paper Published — ECML PKDD 2024
Our paper, “Achieving Counterfactual Explanation for Sequence Anomaly Detection,” presented a framework for identifying minimal changes that transform an anomalous sequence into a normal one. Paper · Code
Paper Published — PAKDD 2024
Our paper, “Backdoor Attack Against One-Class Sequential Anomaly Detection Models,” studied clean-label backdoor attacks against one-class sequential anomaly-detection models. Paper · Code
Doctoral Forum Travel Award — SDM 2024
Received the $1,000 SDM’24 Doctoral Forum Travel Award.
Ph.D. Proposal Defense — Passed
Successfully defended my doctoral research proposal on interpretable and robust deep anomaly detection, covering anomaly detection, model interpretability, and robustness against backdoor attacks.
Paper Published — IJCNN 2023
Our paper, “Explainable Sequential Anomaly Detection via Prototypes,” developed a prototype-based framework for explaining sequential anomaly-detection results. Paper · Code