{
  "annotation_id": "9d312548-457c-4d6b-8def-f262f07248b1",
  "date_created": "2026-02-17T05:53:16.174000Z",
  "date_modified": "2026-02-17T05:53:16.174000Z",
  "file_hash": "3f9c01b95ec858e54941cec15b265b6ff09dfc2d829bdc9e37bcab74b8303a91",
  "private": false,
  "record": {
    "abstract": "This research analyzed and compared the multi-modal approach in the Vision Transformer(EVA-ViT) based image encoder with the LlaMA or ChatGPT LLM to reduce the hallucination problem and detect diseases in chest x-ray images. In this research, we utilized the NIH Chest X-ray image to train the model and compared it in image-based RAG, text-based RAG, and baseline. [3] [5] In a result, the text-based RAG[2] e!ectively reduces the hallucination problem by using external knowledge information, and the image-based RAG improved the prediction con\"dence and calibration by using the KNN methods. [4] Moreover, the GPT LLM showed better performance, a low hallucination rate, and better Expected Calibration Error(ECE) than Llama Llama-based model. This research shows the challenge of data imbalance, a complex multi-stage structure, but suggests a large experience environment and a balanced example of use.",
    "arxiv_id": "2601.08226",
    "authors": [
      "Alexander Shim",
      "Khalil Saieh",
      "Samuel Clarke"
    ],
    "categories": [
      "cs.CV",
      "cs.AI"
    ],
    "license": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
    "title": "Knowledge-based learning in Text-RAG and Image-RAG",
    "url": "https://arxiv.org/abs/2601.08226",
    "version": "v1"
  },
  "schema_id": "dorsal/arxiv",
  "source": {
    "execution_id": "b389ac6e-1fc5-4226-9517-e8ee6dfa6b70",
    "id": "arXiv Dataset",
    "type": "Model",
    "variant": "snapshot-2026-01-17",
    "version": "0.1.0"
  },
  "user_id": 1000002
}