VALOR-BENCH

VALOR-BENCH is a comprehensive human-annotated dataset covering hallucinations in large vision-language models, with a focus on measuring hallucinations in generative tasks.

Data and Resources

Cite this as

Haoyi Qiu, Wenbo Hu, Zi-Yi Dou, Nanyun Peng (2024). Dataset: VALOR-BENCH. https://doi.org/10.57702/2krv5ua9

DOI retrieved: December 16, 2024

Additional Info

Field Value
Created December 16, 2024
Last update December 16, 2024
Defined In https://doi.org/10.48550/arXiv.2404.13874
Author Haoyi Qiu
More Authors
Wenbo Hu
Zi-Yi Dou
Nanyun Peng
Homepage https://github.com/haoyiq114/VALOR