Publications

Preprint An LLM-Based Agent-Oriented Approach for Automated Code Design Issue Localization

Preprint CAPS: Code Abstraction-Based Pre-training Strategy for Smaller Language Models for Code

Preprint NEURALSTATE: An Approach for Detecting Protocol Violations in a Deep Learning Program

ICSE Decomposing a Recurrent Neural Network into Modules for Enabling Reusability and Replacement