Xsum, WritingPrompts, and SQuAD
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
Xsum, WritingPrompts, and SQuAD Aggregated (test)
99.46GPT2-XL
15
Xsum, WritingPrompts, and SQuAD generated by GPT-5-Chat (test)
81.33AUROC
15
Xsum, WritingPrompts, and SQuAD generated by GPT-4.1-mini (test)
80.25AUROC
15
Xsum, WritingPrompts, and SQuAD Gemini-1.5-Flash (test)
75.14AUROC
15