Md Rayhanul Masud
PhD in Computer Science, University of California, Riverside.
Machine learning, software engineering, and cybersecurity for open-source ecosystems.
About Me
I recently completed my Ph.D. in the Department of Computer Science at the University of California, Riverside (UCR), where I was fortunate to be advised by Prof. Michalis Faloutsos. My doctoral research sits at the intersection of machine learning, software engineering, and cybersecurity, where I built ML-driven systems that make open-source communities safer and more transparent by analyzing developer behavior, code similarity, and how software evolves.
Across my Ph.D. I designed systems for repository search (MetaSim, RepoScope), cross-platform identity linkage (GeekMAN), and malware ecosystem analysis (MAGNET), with publications at venues including WWW, ICSME, ASONAM, and SNAM. Before my Ph.D., I spent two years as a software engineer building production systems, which shapes how I approach machine learning today.
My research is in applied machine learning, with a recent focus on the safety and reliability of language models. My recent works include reward engineering for RL on software tasks (Findings of EMNLP 2026), TAME for diagnosing emergent misalignment (EMNLP UncertaiNLP Workshop 2026), and an item-level audit of bias steering (NeurIPS MusIML Workshop 2026).
I earned my B.Sc. in Computer Science and Engineering from Bangladesh University of Engineering and Technology (BUET).
My current research interests span machine learning, large language models, reinforcement learning, alignment, post-training, and model evaluation. I am currently seeking postdoctoral and ML/AI industry positions where rigorous, production-minded work drives real-world impact.
Recent News
-
2026Our paper, "Higher BBQ Scores, More Unnecessary Abstention: An Item-Level Audit of Bias Steering," was accepted to NeurIPS MusIML Workshop 2026.
-
2026Our paper, "TAME: Token Attribution and Masking for Emergent Misalignment," was accepted to EMNLP UncertaiNLP Workshop 2026.
-
2026Our paper, "Hacking Vehicles with LLMs: Generating and Detecting Attacks in Automotive Networks," was accepted to CAINE 2026.
-
Jul 2026Our paper, "Reward Engineering for Software Tasks: A Survey of Reinforcement Learning Approaches," was accepted to EMNLP 2026 (Findings).
-
Jul 2026Our paper, "MAGNET: A software-engineering view of the malware ecosystem on the GitHub network," was accepted to ASONAM 2026.
-
Jun 2026Completed my Ph.D. and graduated from UC Riverside.
-
Jan 2026Received the Dissertation Completion Fellowship from UC Riverside.
-
May 2025Passed my dissertation proposal exam.
-
Nov 2024Presented "MetaSim" at the CRESP Industry Day at UCR.
-
Oct 2024Presented our paper, "MetaSim," at ICSME 2024.
-
Aug 2024Our paper, "MetaSim: A search engine for finding similar GitHub repositories," was accepted to ICSME 2024.
-
Jul 2024Our paper, "Disambiguating usernames across platforms: the GeekMAN approach," was accepted to Social Network Analysis and Mining (SNAM).
-
May 2024Presented our work, "Who is Creating Malware Repositories on GitHub and Why?," at The Web Conference (WWW) 2024.
-
Mar 2024Our paper, "Who is Creating Malware Repositories on GitHub and Why?," was accepted to The Web Conference (WWW) 2024.
-
Nov 2023Presented our paper "GeekMAN: Geek-oriented username Matching Across online Networks" at ASONAM 2023.
-
Jul 2023Presented our poster on "Educational but Malicious Repositories" at KDD 2023.
-
Sep 2021Started my Ph.D. at UC Riverside.