Scholar
Nikunj Saunshi
Google Scholar ID: F24vXggAAAAJ
Research Scientist, Google
Citations & Impact
Citations
2,715
H-index
16
i10-index
17
Publications
20
Co-authors
56
list available
Contact
Publications
Resume (English only)
Academic Achievements
- Paper 'Reasoning with Latent Thoughts: On the Power of Looped Transformers' accepted at ICLR 2025.
- Paper 'On the Inductive Bias of Stacking Towards Improving Reasoning' accepted at NeurIPS 2024.
- Paper 'Can Looped Transformers Learn to Implement Multi-step Gradient Descent for In-context Learning?' accepted at ICML 2024.
- Paper 'Task-Specific Skill Localization in Fine-tuned Language Models' accepted at ICML 2023.
- Paper 'Understanding Influence Functions and Datamodels via Harmonic Analysis' presented at ICLR 2023.
- Paper 'New Definitions and Evaluations for Saliency Methods: Staying Intrinsic, Complete and Sound' accepted at NeurIPS 2022 (Oral).
- Selected as Notable Reviewer for ICLR 2023.
