Jiantang Huang

Hi, I’m Jiantang Huang — a research assistant in Prof. Hang Jiang’s Human-AI Nexus Group, working on LLM agents and reasoning. I hold an MS in Computer and Information Science from Northeastern University and a B.S. in Financial Mathematics (with a CS minor) from The Ohio State University.

My current interests center on agent-to-agent communication and coordination — how multiple LLM agents negotiate, share context, and reach joint decisions — alongside abstract reasoning (ARC-AGI-style generalization) and world models for grounding agent behavior. Recent work includes chain-of-thought distillation for instruction-tuned LLMs (SFT vs. DPO/ORPO on Flan-T5), a Gemini-powered AI dance coaching system using RAG, and an experimental framework for analyzing social-boundary failures in LLM-based agents. I also published research on slow-motion video synthesis for basketball using frame interpolation (arXiv:2511.11644).

I work in Python, C/C++, Go, Java, and Bash, and routinely use TensorFlow, PyTorch, OpenCV, MLflow, Docker, and Azure. I’m Microsoft Certified in Azure AI Fundamentals.

Feel free to reach out by email or LinkedIn — I’m always happy to discuss research, collaborations, or new opportunities.