datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
wisdombench
WisdomBench and Wisdom Science Data
WisdomBench is a longitudinal benchmark for measuring whether an AI agent changes after repeated exposure to feedback and failure.
This dataset contains 3,600 scored evaluation events under the included conditions (3 models x 4 strategies x 20 tasks x 5 rounds x 3 seeds).
It also mirrors the Wisdom Science Research Portfolio release:
Zenodo record: https://zenodo.org/records/20027295
Portfolio DOI: 10.5281/zenodo.20027295
Portfolio folder:… See the full description on the dataset page: https://huggingface.co/datasets/MMJBDS/wisdombench.Quilt_VQA
Dataset Card for "Quilt_VQA"
Paper: Quilt-LLaVA: Visual Instruction Tuning by Extracting Localized Narratives from Open-Source Histopathology Videos
Paper or resources for more information:
https://quilt-llava.github.io/
Description and Details
To evaluate Quilt-LLaVA, alongside public VQA pathology datasets, we also generated Quilt-VQA by extracting Q&A dataset from naturally occurring questions/answers given in the videos. With the help of GPT4 and some handcrafted… See the full description on the dataset page: https://huggingface.co/datasets/wisdomik/Quilt_VQA.MedicalNarrativesMulti-modal models are data hungry. While datasets with natural images are abundant, medical image datasets can not afford the same luxury.
To enable representation learning for medical images at scale, we turn to YouTube, a platform with a large reservoir of open-source medical pedagogical videos.
We curate MedicalNarratives, a dataset 4.7M medical image-text pairs, with 1M samples containing dense annotations in the form of traces and bounding boxes.
Similar to \textit{think-aloud} studies… See the full description on the dataset page: https://huggingface.co/datasets/wisdomik/MedicalNarratives.QuiltVQA_REDDataset Card for "QuiltVQA_ALL"
Human Generated VQA Dataset for Evaluation
Quilt-VQA
is generated by extracting Q&A dataset from naturally occurring questions/answers given in educational histopathology videos. With the help of GPT4 and some handcrafted algorithms, we collect a rich evaluation dataset of 1283 Q&A pairs.
Top two rows show image-dependent Q&A pairs and bottom two rows show general-knowledge Q&A pairs. The original question posed by the narrator of the video is highlighted… See the full description on the dataset page: https://huggingface.co/datasets/wisdomik/QuiltVQA_RED.