AVATAR: Reinforcement Learning to See, Hear, and Reason Over Video
Yogesh Kulkarni, Pooyan Fazli
arXiv, 2025
project page / arXiv

This paper introduces AVATAR, a framework that improves multimodal reasoning by addressing limitations in standard reinforcement learning. It uses an off-policy architecture to improve data efficiency and introduces Temporal Advantage Shaping (TAS), a novel credit assignment strategy to focus learning on critical reasoning steps, achieving significant gains on audio-visual benchmarks.

ReGATE: Learning Faster and Better with Fewer Tokens in MLLMs
Chaoyu Li, Yogesh Kulkarni, Pooyan Fazli
arXiv, 2025
arXiv

This paper proposes ReGATE (Reference-Guided Adaptive Token Elision), a method to accelerate MLLM training by selectively processing crucial tokens. It uses a teacher-student framework to dynamically identify and bypass less informative tokens, achieving state-of-the-art results on benchmarks like MVBench up to 2x faster and with significantly fewer tokens.

VideoSAVi: Self-Aligned Video Language Models without Human Supervision
Yogesh Kulkarni, Pooyan Fazli
Conference on Language Modeling (COLM 2025)
project page / arXiv

This paper presents VideoSAVi, a method to teach video models better spatial and temporal reasoning without needing human supervision. It works by having the model critique its own reasoning errors to automatically create preference data for training, achieving strong results on benchmarks efficiently.

VideoPASTA: 7K Preference Pairs That Matter for Video-LLM Alignment
Yogesh Kulkarni, Pooyan Fazli
arXiv, 2025
project page / arXiv

This paper introduces VideoPASTA, which improves video models by training them with specially crafted "bad examples" (adversarial preference pairs) that target common errors in spatial, temporal, and cross-frame understanding. It shows this targeted approach is highly efficient, achieving significant performance gains using only 7k preference pairs.

EnsembleNTLDetect: An Intelligent Framework for Electricity Theft Detection in Smart Grid
Yogesh Kulkarni, Sayf Hussain Z, Krithi Ramamritham, Nivethitha Somu
ICDM (workshops), 2021
DOI

This paper introduces "EnsembleNTLDetect," a robust framework designed to detect electricity theft in smart grids using consumer energy data. It uses a combination of techniques to handle missing data, data imbalance, and high dimensionality, employing an ensemble machine learning model to achieve high accuracy in identifying theft patterns compared to existing methods.

Kryptonite: An Adversarial Attack Using Regional Focus
Yogesh Kulkarni, Krisha Bhambani
ACNS, 2021
Springer

This paper proposes "Kryptonite," an efficient adversarial attack that fools image classifiers (especially for medical images) by adding tiny, hard-to-see noise mainly to the most important part of the image (Region of Interest). It causes significant misclassification with less image distortion compared to other methods.

Intensive Image Malware Analysis and Least Significant Bit Matching Steganalysis
Yogesh Kulkarni, Anurag Gorkar
IEEE Big Data, 2020
DOI

This paper introduces "AnImAYoung," a framework to detect malware hidden within images using various methods like embedding code in metadata (EXIF) or using steganography (LSB Matching). It analyzes image data/metadata for suspicious code and uses an efficient machine learning ensemble to detect hidden data in image pixels. The system is designed to be fast and accurate, making it suitable for analyzing large volumes of image data.