Research output per year
Research output per year
Personal profile
Biography
Utsav is a PhD Candidate at Macquarie University working on Mechanistic Interpretability of Language Models and its Safety. He previously completed his Postgraduate Research (MRes), with his thesis focusing on mitigating the side effects of safety mechanisms, particularly over-refusal.
His current interest lies in correlating the internal representation of LLMs with safety features; with a broader goal of improving reliability of language models.
Fingerprint
Dive into the research topics where Utsav Maskey is active. These topic labels come from the works of this person. Together they form a unique fingerprint.
- 1 Similar Profiles
Collaborations and top research areas from the last five years
Recent external collaboration on country/territory level. Dive into details by clicking on the dots or
-
MaiBERT: A pre-training corpus and language model for low-resourced Maithili language
Yadav, S., Yadav, R. K., Maskey, U., Kashyap, G. S., Gautam, G. & Naseem, U., 2026, LoResLM 2026: Proceedings of the Second Workshop on Language Models for Low-Resource Languages. Hettiarachchi, H., Ranasinghe, T., Plum, A., Rayson, P., Mitkov, R., Gaber, M., Premasiri, D., Tan, F. A. & Uyangodage, L. (eds.). Kerrville, TX: Association for Computational Linguistics (ACL), p. 444-452 9 p.Research output: Chapter in Book/Report/Conference proceeding › Conference proceeding contribution › peer-review
Open Access1 Link opens in a new tab Citation (Scopus) -
SafeConstellations: mitigating over-refusals in LLMs through task-aware representation steering
Maskey, U. A. Y., Yadav, S., Dras, M. & Naseem, U., 2026, The 64th Annual Meeting of the Association for Computational Linguistics (ACL 2026): proceedings of the Conference - volume 1: long papers. Kerrville, TX: Association for Computational Linguistics, p. 44424-44440 17 p.Research output: Chapter in Book/Report/Conference proceeding › Conference proceeding contribution › peer-review
Open AccessFile -
Should LLM safety be more than refusing harmful instructions?
Maskey, U., Dras, M. & Naseem, U., Aug 2026, In: Machine Learning. 115, 8, p. 1-28 28 p., 186.Research output: Contribution to journal › Article › peer-review
Open AccessFile -
ALTA 2025 Tutorial: welcome letter: Alignment of Large Language Models with Human Preferences and Values
Naseem, U. A. K., Kashyap, G. S., Ren, K., Zhang, Y., Maskey, U., Nadeem, A. & Ren, J., Nov 2025, Proceedings of the 23rd Annual Workshop of the Australasian Language Technology Association. Kerrville, TX: Association for Computational Linguistics, p. 245 1 p.Research output: Chapter in Book/Report/Conference proceeding › Foreword/postscript/introduction › peer-review
Open AccessFile -
Benchmarking Large Language Models for cryptanalysis and side-channel vulnerabilities
Maskey, U., Zhu, C. & Naseem, U., 2025, EMNLP 2025: The 2025 Conference on Empirical Methods in Natural Language processing. Kerrville, TX: Association for Computational Linguistics, p. 19849-19865 17 p.Research output: Chapter in Book/Report/Conference proceeding › Conference proceeding contribution › peer-review
Open AccessFile3 Link opens in a new tab Citations (Scopus)