Pinned Loading

1. grpo-verified-reasoner grpo-verified-reasoner Public

   Post-training Qwen3-4B using SFT + GRPO with verifiable code execution rewards (RLVR).

   Jupyter Notebook 1

2. summa_technologica summa_technologica Public

   A multi-agent AI brainstorming tool that applies the Scholastic method of structured disputation to modern questions.

   Python

3. geodesic geodesic Public

   Interactive visualization of null geodesics in the Kerr metric using WebGL ray marching.

   HTML

4. multimodal-xray-agent multimodal-xray-agent Public

   A deployable, multi-agent system that mimics a radiologist’s diagnostic workflow using cutting-edge vision-language models (VLMs), cross-modal retrieval, and structured agent collaboration. Built w…

   Jupyter Notebook 1 1

5. llm-finetuning-summarizer llm-finetuning-summarizer Public

   A full-stack domain-specialized QA system built on LoRA-fine-tuned Mistral-7B-Instruct-v0.3, trained on curated papers in LLM fine-tuning. Features hybrid RAG, semantic vector search with FAISS, La…

   Jupyter Notebook

6. swinunet-medical-segmentation swinunet-medical-segmentation Public

   End-to-end deep learning pipeline for semantic segmentation of skin lesions using the ISIC 2018 Challenge Dataset.

   Jupyter Notebook 1