RxLM-Med: A multimodal clinical AI agent featuring System 2 reasoning, cross-lingual hierarchical RAG (BM25 + FAISS + RRF), deterministic medical calculation engine, and Traffic Light Protocol (TLP) safety alignment — built on Qwen-VL with LoRA fine-tuning, SFT/DPO alignment, and INT4 quantization for real-world lab report interpretation.

Experiments on System 2 reasoning — neuro-symbolic learning, inference-time scaling, LLM agents, RL post-training in LLMs, and Graph-based retrieval.