curriculum vitae

Undergraduate Researcher, CLIP Lab

Advisors: Rachel Rudinger, Jordan Boyd-Graber, & Mohit Iyyer

• AI agent feedback for long-form text generation
• Calibrating human reliance and susceptibility to VQA model explanations
• Artifacts and distractor quality in human-written vs. LLM-generated MCQs
• Adversarial benchmarking of VLMs on out-of-context video misinformation
• Causal interpretability of syntactic learning in data-constrained LMs

Visiting Researcher, ACL2 Lab & NSF

Advisor: Kenny Zhu

• Unsupervised techniques for feline language discovery and categorization
• AniVoice-cat: 26,000+ annotated cat vocalizations and 57 discovered phones

Member of Technical Staff (San Francisco, CA)

• Ran HackAPrompt, the world's largest AI red-teaming hackathon
• Robustness of AI safety judges to universal and multimodal CBRNE jailbreaks

Researcher, FIRE Sustainability Analytics Lab

Advisor: Thanicha Ruangmas

• Climate policy frameworks based on automated OCR of 50,000+ 10-K filings

Researcher, Reinforcement Learning at Brown Group

Advisor: Michael Littman

• Human-readable reward interfaces letting non-experts specify RL tasks

Under Submission

1. Harder Benchmarks Are Not Always Better: Evaluating LLM Distractors for Multiple-Choice Benchmarks
   Under review at ACL Rolling Review (ARR)
   Atrey Desai, Nishant Balepur & Rachel Rudinger.
2. Asking About Occluded Content Marks Adversarial Multimodal Questions
   Under review at ACL Rolling Review (ARR)
   HyoJung Han*, Atrey Desai*, Ishan Upadhyay, Tianyi Zheng, Eve Fleisig, Aparna Garimella, Marine Carpuat & Jordan Boyd-Graber.

Peer-reviewed Proceedings

1. A Preview of Computational Animal Linguistics.
   ACM Computing Surveys
   Tuan Dang*, Atrey Desai*, Hridayesh Lekhak*, Tirza Panunto*, Lindsay Pike*, Theron Wang* & Kenny Zhu*.
2. Filling in the Mechanisms: How do LMs Learn Filler-Gap Dependencies under Developmental Constraints?
   ACL 2026 (Findings) · Oral at TLS 2026
   Atrey Desai & Sathvik Nair.
3. BenchMarker: An Education-Inspired Toolkit for Highlighting Flaws in Multiple-Choice Benchmarks.
   ACL 2026
   Nishant Balepur, Bhavya Rajasekaran, Jane Oh, Michael Xie, Atrey Desai, …, Jordan Boyd-Graber.
4. Test-Time Reasoners Are Strategic Multiple-Choice Test-Takers.
   ACL 2026
   Nishant Balepur, Atrey Desai & Rachel Rudinger.
5. Language Models Generate Multiple-Choice Questions with Artifacts.
   MASC-SLL 2025
   Atrey Desai, Nishant Balepur & Rachel Rudinger.
6. Reinforcement Learning As End-User Trigger-Action Programming.
   IML Workshop @ AAAI 2022 · RLDM 2022
   Chace Hayhurst, Hyojae Park, Atrey Desai, Suheidy De Los Santos & Michael Littman.

Filler-Gap Dependencies under Developmental Constraints in LMs

• Texas Linguistics Society (Feb 2026)
• UMD Computational Cognitive Science Reading Group (Feb 2026)

Adaptor Grammars and Neural Networks for Feline Lexical Discovery

• University of Maryland (Nov 2024)
• University of Texas at Arlington (Jul 2024)

Elected by 4,200+ CS undergraduates (nation's largest student body) and appointed by faculty.

Represented 1000+ peers, ran events & workshops, and reformed class curricula.

Mentoring

• REU CAAR: Nathan Nguyen, Nghia Truong, Yasmine Tajeddin (2026)
• Office of Undergraduate Research: Juan Cortés, Kemisola Benson, Vivian Akpala (2025)
• Technica: Savya Miriyala, Tanya Grover, Jessica Ononye, Nakshatra Hiray (2024)
• Robotics Curriculum: designed and led for MSET (2020 – 2022)