Improve dataset card: Add metadata and comprehensive content

README.md

@@ -1,4 +1,16 @@
 ---
+task_categories:
+- image-text-to-text
+license: cc-by-4.0
+language:
+- en
+library_name: datasets
+tags:
+- medical
+- multimodal
+- in-context-learning
+- benchmark
+- vqa
 dataset_info:
   features:
   - name: answer
@@ -45,3 +57,49 @@ configs:
   - split: train
     path: data/train-*
 ---
+
+# SMMILE: An Expert-Driven Benchmark for Multimodal Medical In-Context Learning
+
+<div align="center">
+  <img src="https://huggingface.co/datasets/smmile/SMMILE/resolve/main/figures/logo_final.png" alt="SMMILE Logo" width="500"/>
+</div>
+
+- **Paper**: [SMMILE: An Expert-Driven Benchmark for Multimodal Medical In-Context Learning](https://huggingface.co/papers/2506.21355)
+- **Project Page**: [https://smmile-benchmark.github.io](https://smmile-benchmark.github.io)
+- **GitHub Repository**: [https://github.com/smmile-benchmark/SMMILE](https://github.com/smmile-benchmark/SMMILE)
+- **Hugging Face Hub**: [https://huggingface.co/datasets/smmile/SMMILE](https://huggingface.co/datasets/smmile/SMMILE)
+
+## Introduction
+
+Multimodal in-context learning (ICL) remains underexplored despite significant potential for domains such as medicine. Clinicians routinely encounter diverse, specialized tasks requiring adaptation from limited examples, such as drawing insights from a few relevant prior cases or considering a constrained set of differential diagnoses. While multimodal large language models (MLLMs) have shown advances in medical visual question answering (VQA), their ability to learn multimodal tasks from context is largely unknown.
+
+We introduce SMMILE, the first expert-driven multimodal ICL benchmark for medical tasks. Eleven medical experts curated problems, each including a multimodal query and multimodal in-context examples as task demonstrations. SMMILE encompasses 111 problems (517 question-image-answer triplets) covering 6 medical specialties and 13 imaging modalities. We further introduce SMMILE++, an augmented variant with 1038 permuted problems. A comprehensive evaluation of 15 MLLMs demonstrates that most models exhibit moderate to poor multimodal ICL ability in medical tasks. In open-ended evaluations, ICL contributes only 8% average improvement over zero-shot on SMMILE and 9.4% on SMMILE++. We observe a susceptibility for irrelevant in-context examples: even a single noisy or irrelevant example can degrade performance by up to 9.5%. Moreover, example ordering exhibits a recency bias, i.e., placing the most relevant example last can lead to substantial performance improvements by up to 71%. Our findings highlight critical limitations and biases in current MLLMs when learning multimodal medical tasks from context.
+
+## Dataset Access
+
+The SMMILE dataset is available on Hugging Face. You can load it using the `datasets` library:
+
+```python
+from datasets import load_dataset
+load_dataset('smmile/SMMILE', token=YOUR_HF_TOKEN)
+load_dataset('smmile/SMMILE-plusplus', token=YOUR_HF_TOKEN)
+```
+
+Note: You need to set your Hugging Face token as an environment variable:
+```bash
+export HF_TOKEN=your_token_here
+```
+
+## Citation
+
+If you find this dataset useful for your research, please cite the corresponding paper:
+
+```bibtex
+@article{rieff2025smmile,
+      title={SMMILE: An Expert-Driven Benchmark for Multimodal Medical In-Context Learning},
+      author={Rieff, Maximilian and Varma, Mayank and Rabow, Oliver and Adithan, Swetha and Kim, Jaehee and Chang, Kyeong and Lee, Han and Rohatgi, Nikhil and Bluethgen, Conrad and Muneer, Mohammed Shaheer and Delbrouck, Jean-Baptiste and Moor, Michael},
+      journal={arXiv preprint arXiv:2506.21355},
+      year={2025},
+      url={https://arxiv.org/abs/2506.21355}
+}
+```