|

Aditya Mehrotra, Alice Johnson

arXiv preprint • 2024

A comprehensive survey of techniques for scaling machine learning systems across distributed computing environments, covering both data and model parallelism approaches.

Aditya Mehrotra, John Smith, Jane Doe

NeurIPS 2024 • 2024

We propose a novel method for parameter-efficient fine-tuning of large language models that reduces memory requirements by 70% while maintaining comparable performance to full fine-tuning.

Bob Williams, Aditya Mehrotra, Carol Davis

CVPR 2023 • 2023

We investigate adversarial attack vectors in state-of-the-art computer vision models and propose a new defense mechanism that improves robustness by 35%.