I am currently an Applied Scientist at AWS. I work on problems in privacy, model architectures, and reasoning in deep neural networks. Previously, I was a PhD student at the Computer Science Department in UCLA, working with Prof. Stefano Soatto in the UCLA Vision Lab. Before that I was an undergrad in Electrical Engineering at the Indian Institute of Technology, Bombay.

Email: adityagolatkar [AT] ucla [dot] edu

Publications & Workshops

• R. Zabounidis, A. Golatkar, M. Kleinman, A. Achille, W. Xia, S. Soatto
  Re-FORC: Adaptive Reward Prediction for Efficient Chain-of-Thought Reasoning
  NeurIPS 2025 Efficient Reasoning Workshop, 2025
  [arXiv]

• A. Achille, A. S. Golatkar, B. H. Bowman, M. Kearns, C. Klingenberg, S. Soatto
  Systems and methods for training machine learning models using emulated datasets
  US Patent App. 18/434,291, 2025
  

• T. Y. Liu, A. Achille, M. Trager, A. Golatkar, L. Zancato, S. Soatto
  PICASO: Permutation-Invariant Context Composition with State Space Models
  International Conference on Learning Representations (ICLR), 2025
  [arXiv]

• E. Nunez, L. Zancato, B. Bowman, A. Golatkar, W. Xia, S. Soatto
  Expansion span: Combining fading memory and retrieval in hybrid state space models
  2nd International Conference on Neuro-symbolic Systems (NeuS), 2024
  [arXiv]

• L. Zancato, A. Seshadri, Y. Dukler, A. S. Golatkar, Y. Shen, B. Bowman, et al.
  B’MOJO: Hybrid State Space Realizations of Foundation Models with Eidetic and Fading Memory
  Advances in Neural Information Processing Systems (NeurIPS) 37, 2024
  [arXiv]

• B. Biggs, A. Seshadri, Y. Zou, A. Jain, A. Golatkar, Y. Xie, A. Achille, et al.
  Diffusion soup: Model merging for text-to-image diffusion models
  European Conference on Computer Vision (ECCV), 2024
  [arXiv]

• A. Golatkar, A. Achille, L. Zancato, Y. X. Wang, A. Swaminathan, S. Soatto
  CPR: Retrieval Augmented Generation for Copyright Protection
  Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024
  [arXiv]

• A. Golatkar, A. Achille, A. Swaminathan, S. Soatto
  Training data protection with compositional diffusion models
  arXiv preprint arXiv:2308.01937, 2023
  [arXiv]

• T. Y. Liu, A. Golatkar, S. Soatto
  Tangent transformers for composition, privacy and removal
  International Conference on Learning Representations (ICLR), 2024
  [arXiv]

• Y. Dukler, B. Bowman, A. Achille, A. Golatkar, A. Swaminathan, S. Soatto
  SAFE: Machine unlearning with shard graphs
  Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 2023
  [arXiv]

• T. Y. Liu, A. Golatkar, S. Soatto, A. Achille
  Integral continual learning along the tangent vector field of tasks
  arXiv preprint arXiv:2211.13108, 2022
  [arXiv]

• M. Rida Rammal, A. Achille, A. Golatkar, S. Diggavi, S. Soatto
  On Leave-One-Out Conditional Mutual Information For Generalization
  Advances in Neural Information Processing Systems (NeurIPS), 2022
  [arXiv]

• A. Golatkar, A. Achille, Y. Wang, A. Roth, M. Kearns, S. Soatto
  Mixed Differential Privacy for Computer Vision
  Conference on Computer Vision and Pattern Recognition (CVPR), 2022, (Oral Presentation)
  [arXiv]

• S. Tsuei, A. Golatkar, S. Soatto
  Scene Uncertainty and the Wellington Posterior of Deterministic Image Classifiers
  arXiv preprint arXiv:2106.13870, 2021
  [arXiv]

• A. Golatkar, A. Achille, M. Polito, A. Ravichandran, S. Soatto
  Mixed-Privacy Forgetting in Deep Networks
  Conference on Computer Vision and Pattern Recognition (CVPR), 2021
  [arXiv]

• A. Achille, A. Golatkar, M. Polito, A. Ravichandran, S. Soatto
  LQF: Linear Quadratic Fine-tuning
  Conference on Computer Vision and Pattern Recognition (CVPR), 2021, (Oral Presentation)
  [arXiv]

• A. Golatkar, A. Achille, S. Soatto
  Forgetting Outside the Box: Scrubbing Deep Networks of Information Accessible from Input-Output Observations
  European Conference on Computer Vision (ECCV), 2020
  [arXiv]

• A. Golatkar, A. Achille, S. Soatto
  Eternal Sunshine of the Spotless Net : Selective forgetting in Deep Networks
  Conference on Computer Vision and Pattern Recognition (CVPR), 2020
  [arXiv]

• A. Golatkar, A. Achille, S. Soatto
  Time Matters in Regularizing Deep Networks: Weight Decay and Data Augmentation Affect Early Learning Dynamics, Matter Little Near Convergence
  Neural Information Processing Systems (NeurIPS), 2019
  [arXiv]

• R. Das ^#, A. Golatkar ^#, S. Awate ( ^# denotes equal contribution)
  Sparse Kernel PCA for Outlier Detection
  IEEE International Conference on Machine Learning and Application (ICMLA), 2018, (Oral Presentation)
  [arXiv]

• A. Golatkar, D. Anand, A. Sethi
  Classification of Breast Cancer Histology Using Deep Learning
  International Conference on Image Analysis and Recognition (ICIAR), 2018
  [arXiv]

• P. Kothari, A. Sadhu, A. Golatkar, R. Das (all authors contributed equally)
  iFood Challenge, FGVC Workshop
  Conference on Computer Vision and Pattern Recognition (CVPR), 2018
  Finished $2^{nd}$ in the public leaderboard and $3^{rd}$ in the private leaderboard (Team name : Invincibles). Leaderboard Link. Invited to present our method at CVPR 2018 (slides can be found here)

Work Experience

• Applied Scientist
  AWS AI
  Summer 2023 - Present

• Applied Scientist Intern
  AWS AI, Pasadena
  Summer 2022

• Applied Scientist Intern
  AWS AI, Pasadena
  Summer 2021

• Applied Scientist Intern
  AWS AI, Pasadena
  Summer 2020

• Graduate Student Researcher
  UCLA Vision Lab
  Jan 2019 - June 2023

• Applied Scientist Intern
  Amazon Go, Seattle
  Summer 2019

• AI Research Intern
  Wadhwani Institute for AI, Mumbai
  Summber 2018

• PRAIRIE Artificial Intelligence Summer School, Grenoble, France
  Organized by Inria and NAVER LABS Europe
  Selected for a poster presentation [Poster]
  Summer 2018

• Quantitative Research Intern
  Nomura Holdings, Mumbai
  Summer 2017

• Software Developement Intern
  Indus OS, Mumbai
  Summer 2016

Awards

• NeurIPS 2019 Travel Award
  

• INSPIRE Scholarship, Government of India
  

• State Scholarship, Government of Maharashtra
  

Talks

• Amazon Machine Learning Conference 2020