Aditya Agrawal

Aditya Agrawal

Carnegie Mellon University - School of Computer Science

Carnegie Mellon University

Biography

I am a first year student at Carnegie Mellon University persuing my Masters of Science in Machine Learning.

I am currently working with Martin Ma at the Russ lab on improving sample efficiency of Large Language Model.

Education

  • Masters of Science in Machine Learning, 2025

    Carnegie Mellon University - School of Computer Science

  • Bachelors of Engineering in Computer Science, 2024

    Birla Institute of Technology and Science, Pilani - KK Birla Goa Campus

Experience

 
 
 
 
 
Research Intern
Adobe MDSR Lab
Oct 2022 – Present Remote

Advisors :

  • Mr, Yaman Kumar

Studying the correlation between Attentivity and Memoribility on video Ads.

 
 
 
 
 
Undergraduate Researcher
APP Center for AI Research, BITS Goa
Jan 2022 – Jan 2023 BITS Pilani, Goa Campus

Advisors :

  • Dr. Ashwin Srinivasan, Dr. Lovekesh Vig, Dr. Tanmay Verlekar, Dr. Tirthraj Dash, Dr. Gautam Shroff

Using semantic loss to train a neural network to extract features for logical theory on time-series data and applying it to ecg arrhythmia detection for explainable predictions.

 
 
 
 
 
Intern
Oyla
May 2022 – Aug 2022 Remote

Advisors:

  • Dr. Raghavendra Singh

Using DPT model on RGB-D images by changing the feature extractors to incorporate depth information in embeddings.

Projects

*
All Computer Vision Reinforcement Learning
Disentangled Predictive Representation for Meta-Reinforcement Learning
Paper Implementation of ‘Disentangled Predictive Representation for Meta-Reinforcement Learning’
Disentangled Predictive Representation for Meta-Reinforcement Learning
Vformer
A modular PyTorch library for vision transformers models
Follow
Vformer

Recent Publications

Martin Q. Ma, Willis Guo, Aditya Agrawal, Ankit Gupta, Paul Pu Liang, Russ Salakhutdinov, Louis-Philippe Morency (2025). Video Active Perception: Efficient Inference-Time Long-Form Video Understanding with Vision-Language Models. At ICLR 2025 (under review).

Cite

Aditya Agrawal, Ashmit Khandelwal, Aanisha Bhattacharyya, Yaman Kumar, Somesh Singh, Uttaran Bhattacharya, Ishita Dasgupta, Stefano Petrangeli, Rajiv Ratn Shah, Changyou Chen, Balaji Krishnamurthy (2024). Large Content And Behavior Models To Understand, Simulate, And Optimize Content And Behavior. At ICLR 2024.

PDF Cite Abstract

Victor Boutin, Rishav Mukherji, Aditya Agrawal, Sabine Muzellec, Thomas Fel, Thomas Serre, Rufin VanRullen (2024). Latent Representation Matters: Human-like Sketches in One-shot Drawing Tasks. At NeurIPS 2024.

PDF Cite Abstract

Vedant Shah, Aditya Agrawal, Lovekesh Vig, Ashwin Srinivasan, Gautam Shroff, Tanmay Verlekar (2022). Neural Feature-Adaptation for Symbolic Predictions Using Pre-Training and Semantic Loss. On ArXiv (preprint).

PDF Cite Abstract

Contact