Paper: [https://huggingface.co/papers/2602.03143](https://huggingface.co/papers/2602.03143)

Training set: [https://huggingface.co/datasets/baohao/sage_train](https://huggingface.co/datasets/baohao/sage_train)

Validation set: [https://huggingface.co/datasets/baohao/sage_validation](https://huggingface.co/datasets/baohao/sage_validation)

Github: [https://github.com/BaohaoLiao/SAGE.git](https://github.com/BaohaoLiao/SAGE.git)