Bhash Abeysinghe

About Me

Bhashithe Abeysinghe is a researcher specializing in agentic AI and multi-agent systems for policy innovation and resilient decision-making. His work focuses on designing frameworks that enable collaborative, adaptive policymaking through AI-driven simulations and coordination strategies. In addition to his expertise in policy-oriented AI, Bhashithe has extensive experience in Natural Language Processing (NLP), graph mining, and machine learning, applying these techniques to extract insights from complex data, model relationships, and build intelligent systems. His research integrates advanced AI architectures, language understanding, and network analysis to tackle real-world governance challenges and enhance the robustness of policy ecosystems.

Learn more about my work here.

Current Research Articles

Below is a list of my recent research articles. For the full list, visit my Google Scholar profile.

• The Challenges of Evaluating LLM Applications: An Analysis of Automated, Human, and LLM-Based Approaches
• Decoding Student Insights: Analyzing Response Change in NAEP Mathematics Constructed Response Items.
• Ask NAEP: A Generative AI Assistant for Querying Assessment Information
• Inferring stances of silent-participants in Twitter chatter using Label Propagation
• Utilizing Multi-modal Weak Signals to Improve User Stance Inference in Social Media
• Misinformation in Social Media Platforms and Web Articles: a Dataset to Infer User Stance
• Unsupervised user stance detection on tweets against web articles using sentence transformers
• POSLAN: disentangling chat with positional and language encoded post embeddings
• An Interactive Lambda Calculus Interpreter and Visualization Tool
• An Interactive Lambda Calculus Execution Engine
• A Semantic Segmentation Approach to Recognize Assault Rifles in ISIS Propaganda Images