AURORABench

Read the paper here: https://arxiv.org/abs/2407.03471. Please check out our GitHub repository if you simply want a json and image folders (I personally find parquet confusing): https://github.com/McGill-NLP/AURORA

Papers using AURORABench (2)

Self-Improving World Modelling with Latent Actions2026EditReward: A Human-Aligned Reward Model for Instruction-Guided Image Editing2025