Projects

Robustness and Security of Vision–Language Models
We investigate how multimodal foundation models can be attacked and defended, characterizing threats that target not just outputs but the internal reasoning of vision–language models.
Robustness and Security of Vision–Language Models
Operational Safety of Autonomous Coding Agents
We study the real-world safety of LLM-based coding agents, characterizing how they fail during everyday development tasks and what safeguards are needed to deploy them responsibly.
Operational Safety of Autonomous Coding Agents
Reasoning and Planning in Large Language Models
We study how large language models reason, aiming to move beyond local, token-by-token decisions toward reliable global planning through structured guidance and reinforcement learning.
Reasoning and Planning in Large Language Models
Long-Term Risks in ML Systems
This project investigates how feedback loops in ML systems can cause long-term, harmful impacts, and develops tools to detect, analyze, and prevent them before deployment.
Long-Term Risks in ML Systems
Design and Architecture of Data Science Pipelines
We study, design, and analyze the DS pipeline architecture consisting stages such as preprocessing, modeling, training, evaluation, etc.
Design and Architecture of Data Science Pipelines
Safety Assurance of Predictive Systems
We built abstractions of ML systems and inferred preconditions to provide assurance in safety-critical predictions.
Safety Assurance of Predictive Systems
Verifying Neural Networks for Individual Fairness
A modular approach to formally verify neural networks. We specified individual property for SMT solver and verified fairness for specific subpopulations.
Verifying Neural Networks for Individual Fairness
Causal Fairness in Machine Learning Pipeline
We used causal reasoning to measure fairness of components and remove them from machine learning pipeline.
Causal Fairness in Machine Learning Pipeline
Fairness Engineering in ML Models
We have studied the software engineering concerns of fairness in real-world machine learning models.
Fairness Engineering in ML Models
ML Repo Dataset from GitHub
This dataset is created by mining 5M Python program snapshots. The code is transformed to AST for static analysis.
ML Repo Dataset from GitHub