Yamini Bansal

Researcher, Google

Joined

October 2017

Names

Yamini Bansal

Emails

****@g.harvard.edu (Confirmed)

,

****@google.com (Confirmed)

Personal Links

Career & Education History

Researcher

Google (google.com)

2022 – Present

PhD student

Harvard University (harvard.edu)

2016 – 2022

Advisors, Relations & Conflicts

PhD Advisor

David Cox

Present

Coauthor

Present

Coauthor

Present

Undergraduate adviser

Bipin Rajendran

2014 – 2016

Expertise

Theory of deep learning

,

generative models

Present

Publications

Beyond Human Data: Scaling Self-Training for Problem-Solving with Language Models
Avi Singh, John D Co-Reyes, Rishabh Agarwal, Ankesh Anand, Piyush Patil, Xavier Garcia, Peter J Liu, James Harrison, Jaehoon Lee, Kelvin Xu, Aaron T Parisi, Abhishek Kumar, Alexander A Alemi, Alex Rizkowsky, Azade Nova, Ben Adlam, Bernd Bohnet, Gamaleldin Fathy Elsayed, Hanie Sedghi, Igor Mordatch et al. (20 additional authors not shown)
- Accepted by TMLR
- Readers: Everyone
Empirical Limitations of the NTK for Understanding Scaling Laws in Deep Learning
Nikhil Vyas, Yamini Bansal, Preetum Nakkiran
- Accepted by TMLR
- Readers: Everyone
The Unreasonable Effectiveness of Few-shot Learning for Machine Translation
Xavier Garcia, Yamini Bansal, Colin Cherry, George Foster, Maxim Krikun, Melvin Johnson, Orhan Firat
- ICML 2023 Poster
- Readers: Everyone
Limitations of the NTK for Understanding Generalization in Deep Learning
Nikhil Vyas, Yamini Bansal, Preetum Nakkiran
- Published: 01 Feb 2023, Last Modified: 13 Feb 2023
- Submitted to ICLR 2023
- Readers: Everyone
Data Scaling Laws in NMT: The Effect of Noise and Architecture
Yamini Bansal, Behrooz Ghorbani, Ankush Garg, Biao Zhang, Colin Cherry, Maxim Krikun, Behnam Neyshabur, Orhan Firat
- Published: 28 Jan 2022, Last Modified: 13 Feb 2023
- ICLR 2022 Submitted
- Readers: Everyone
Distributional Generalization: Structure Beyond Test Error
Preetum Nakkiran, Yamini Bansal
- Published: 28 Jan 2022, Last Modified: 13 Feb 2023
- ICLR 2022 Submitted
- Readers: Everyone
Distributional Generalization: Characterizing Classifiers Beyond Test Error
Preetum Nakkiran, Yamini Bansal
- 21 May 2021 (modified: 05 May 2023)
- NeurIPS 2021 Submitted
- Readers: Everyone
Revisiting Model Stitching to Compare Neural Representations
Yamini Bansal, Preetum Nakkiran, Boaz Barak
- Published: 09 Nov 2021, Last Modified: 26 May 2025
- NeurIPS 2021 Poster
- Readers: Everyone
Improving the Reconstruction of Disentangled Representation Learners via Multi-Stage Modelling
Akash Srivastava, Yamini Bansal, Yukun Ding, Cole Lincoln Hurwitz, Kai Xu, Bernhard Egger, Prasanna Sattigeri, Joshua B. Tenenbaum, Dan Gutfreund
- 28 Sept 2020 (modified: 12 Oct 2025)
- ICLR 2021 Conference Blind Submission
- Readers: Everyone
Distributional Generalization: A New Kind of Generalization
Preetum Nakkiran, Yamini Bansal
- 28 Sept 2020 (modified: 12 Oct 2025)
- ICLR 2021 Conference Blind Submission
- Readers: Everyone

View all 11 publications

Co-Authors

View all 59 co-authors