Clara Mohri

I am a PhD student at Harvard University, where I am fortunate to be advised by Sham Kakade. I am very happy to be supported by the Kempner Graduate Fellowship.

My research interests include efficient algorithms for training and deploying language models.

Please feel free to reach out to me at cmohri [at] g.harvard.edu.

Publications

Fast Inference via Hierarchical Speculative Decoding

C. Mohri, H. Kaplan, T. Schuster, Y. Mansour, A. Globerson

arXiv preprint, 2025
Budgeted Multiple-Expert Deferral

G. DeSalvo, C. Mohri, M. Mohri, Y. Zhong

arXiv preprint, 2025
Mixture of Parrots: Experts Improve Memorization More Than Reasoning

S. Jelassi, C. Mohri, D. Brandfonbrener, A. Gu, N. Vyas, N. Anand, et al.

International Conference on Learning Representations (ICLR), 2025
Using Language to Extend to Unseen Domains

L. Dunlap, C. Mohri, D. Guillory, H. Zhang, T. Darrell, J. E. Gonzalez

International Conference on Learning Representations (ICLR), 2023
Partial Matrix Completion

E. Hazan, A. T. Kalai, V. Kanade, C. Mohri, Y. J. Sun

Advances in Neural Information Processing Systems (NeurIPS), 2023