Bio
I work as a Research Scientist at Google Research. I defended my Ph.D. in Computer Science at the University of Maryland, advised by Prof. Feizi and Prof. Hajiaghayi. My research studies generative models from two perspectives: data, through pretraining data selection and machine unlearning, and model, through knowledge localization and model editing. I have also proposed methods to monetize large language models by integrating advertisements into their outputs.
More about me
Outside of research, I’m usually on the soccer pitch, building LEGO and other puzzles, or thinking about combinatorics problems. I also keep a close eye on global politics and social trends.
News
[Oct 26] I joined Google Research as a Research Scientist!
[Sep 26] Our paper "SpecHop: Continuous Speculation for Accelerating Multi-Hop Retrieval Agents" is accepted to NeurIPS 2026.
[Jun 26] Maryland Today covered our work on machine unlearning!
[Jun 26] I'll be joining Susquehanna International Group for Summer 2026, let's do some Finance!
[Jan 26] Our paper "Revisiting the Past: Data Unlearning with Model State History" is accepted to ICLR 2026.
[Sep 25] Our paper "Localizing Knowledge in Diffusion Transformers" is accepted to NeurIPS 2025.
[May 25] I'll be joining Adobe as a Research Intern for Summer 2025, working with the Document Intelligence Lab!
[Mar 25] Our paper "RESTOR: Knowledge Recovery via Machine Unlearning," has been accepted to Transactions on Machine Learning Research (TMLR) 2025.
[Mar 25] Our paper "Online Advertisements with LLMs: Opportunities and Challenges" is accepted to ACM SIGecom Exchanges 2025.
[Jan 25] I'll be joining Google as Student Researcher for Spring 2025 to work on data selection in large language model pretraining.
[Nov 24]
See our blogpost on Mechanstic Interpretability Beyond Language Models!
[Sep 24] Our paper "Ad Auctions for LLMs via Retrieval Augmented Generation" is accepted to NeurIPS 2024.
[May 24] Our paper "On Mechanistic Knowledge Localization in Text-to-Image Generative Models" is accepted to ICML 2024.
[Mar 24] I will be joining MOSAIC at Allen Institute for AI (Ai2) as Research Intern in summer 2024!
[Jan 24] Our paper "PRIME: Prioritizing Interpretability in Failure Mode Extraction" is accepted to ICLR 2024.
[Jan 24] Our paper "Robustness of AI-Image Detectors: Fundamental Limits and Practical Attacks" is accepted to ICLR 2024.
[Dec 23] Our paper "Regret Analysis of Repeated Delegated Choice" is accepted to AAAI 2024.
[Nov 23] We posted our draft "Online Advertisements with LLMs: Opportunities and Challenges" on arXiv.
[Oct 23] Our work of Robustness of AI-Image Detectors is covered by
Register,
Wired, and
Arstechnica.
[May 23] Our team, UMD RED, ranked 3-rd in ICPC NAC 2023 and proceeded to ICPC World Finals 2023.
[May 23] Our paper "Delegating to Multiple Agents" is accepted to EC 2023.
[Apr 23] Our paper "Run-off Election: Improved Provable Defense against Data Poisoning Attacks" is accepted to ICML 2023.
[Apr 23] Our paper "Text-To-Concept (and Back) via Cross-Model Alignment" is accepted to ICML 2023.
[Feb 23] Our team, UMD RED, is qualified to compete in ICPC NAC 2023 at UCF.
Affiliation & Research
Google Research (Research Scientist)
10/26 – Now
Reliable AI Lab, University of Maryland (Research Assistant)
09/22 – 09/26
Google Research (Student Researcher) — pretraining data selection
02/25 – 04/25
Susquehanna (SIG) (ML Research Intern) — data selection for intra-day price prediction
06/26 – 08/26
Allen Institute for AI (Ai2) (Research Intern) — data-level machine unlearning
05/24 – 10/24
Adobe Research (Research Intern) — text-to-diagram scientific visualization
05/25 – 08/25
TML Lab, EPFL (Research Intern) — SGD convergence rate
07/21 – 09/21
Publications
* denotes equal contribution
·
α, β denotes alphabetical order of authorship
Powered by Jekyll and Minimal Light theme, with help from 
Claude Code and 
Antigravity.