Anjali Kantharuban

PhD Student, Language Technologies Institute, Carnegie Mellon University

Multilingual NLP · Personalization · User Simulation & Evaluation

Email CV Google Scholar GitHub LinkedIn X

About

I'm a PhD student in the Language Technologies Institute at Carnegie Mellon University, advised by Graham Neubig, graduating in May 2028. My research is about making NLP work for the way people actually talk — low-resource languages and non-standard dialects — and, increasingly, about personalization: how systems should adapt to individual users, and how to evaluate that with realistic user simulation rather than static benchmarks. My research has been supported by an NSF Graduate Research Fellowship.

I'm currently a Research Scientist Intern at Meta Superintelligence Labs, working on multilingual personalization. Before that, I spent a summer at Handshake AI Research building calibrated user simulators for benchmarking multi-turn agents.

Before CMU, I completed an MPhil in Theoretical and Applied Linguistics at the University of Cambridge as a Gates Cambridge Scholar, working with Anna Korhonen and Ivan Vulić on dialect gaps in machine translation and speech recognition, and a BA in Computer Science and Linguistics at UC Berkeley, where I worked with Dan Klein and with Jacob Andreas at MIT. I've studied Tamil and Telugu for five semesters.

News

Oct 2026 CUEing User Simulators is out on arXiv — code and model released.
Sep 2026 Blog post on The Tokenizer Tax released in collaboration with CLEAR.
Aug 2026 Started as a Research Scientist Intern at Meta Superintelligence Labs.
Mar 2026 Co-organizing the AMIYA shared task on Arabic dialect modeling at VarDial 2026.

Experience

Meta — Research Scientist Intern, Superintelligence Labs (Aug 2026 – Jan 2027)
Multilingual personalization on the response quality team.

Handshake AI Research — Research Intern (May – Aug 2026)
Calibrated, human-aligned user simulation for multi-turn agent benchmarking.

Selected Publications

CUEing User Simulators: Calibrated User Embeddings for Multi-Turn Benchmarking
Anjali Kantharuban, Jonas Mueller · Preprint
Paper · Code · Model

IdioleX: Unified and Continuous Representations for Idiolectal and Stylistic Variation
Anjali Kantharuban, Aarohi Srivastava, Fahim Faisal, Orevaoghene Ahia, Antonios Anastasopoulos, David Chiang, Yulia Tsvetkov, Graham Neubig · Preprint
Paper · Code

Stereotype or Personalization? User Identity Biases Chatbot Recommendations
Anjali Kantharuban*, Jeremiah Milbauer*, Maarten Sap, Emma Strubell, Graham Neubig · Findings of ACL, 2025
Paper · Code

Pangea: A Fully Open Multilingual Multimodal LLM for 39 Languages
Xiang Yue*, Yueqi Song*, Akari Asai, Seungone Kim, Jean de Dieu Nyandwi, Simran Khanuja, Anjali Kantharuban, Lintang Sutawika, Sathyanarayanan Ramamoorthy, Graham Neubig · ICLR, 2025
Paper · Code · Model

Quantifying the Dialect Gap and its Correlates Across Languages
Anjali Kantharuban, Ivan Vulić, Anna Korhonen · Findings of EMNLP, 2023
Paper

Counting the Bugs in ChatGPT's Wugs: A Multilingual Investigation into the Morphological Capabilities of a Large Language Model
Leonie Weissweiler, Valentin Hofmann, Anjali Kantharuban, Anna Cai, Ritam Dutt, Amey Hengle, Anubha Kabra, Atharva Kulkarni, Abhishek Vijayakumar, Haofei Yu, Hinrich Schütze, Kemal Oflazer, David R. Mortensen · EMNLP, 2023
Paper · Code

Teaching & Service

Program Committee, VarDial 2026
Co-organizer, AMIYA Shared Task: Arabic Modeling In Your Accent, VarDial 2026
Teaching, Subword Modeling, Carnegie Mellon University (2024)
Head Teaching Assistant, CS61B Data Structures, UC Berkeley (2019–2022)
Outstanding Graduate Student Instructor Award, UC Berkeley (2022)

Contact

The best way to reach me is email: anjaliruban AT cmu DOT edu. I'm also on LinkedIn and X.