Story
Benchmarks
TaskDatasetSOTA resultTrendResults
Story
38.6Semantic Diversity
20
Story
8.36Diversity
19
Story
46.7Exact Match (EM)
14
Story
0.207WD
12
Story (test)
0.96Diversity (DIV)
12
Story
0.998Rewrite AUC (Claude 3.5)
11
Story dataset GPT-5-mini K=5
0.44WD
6
Story
63.6Coherence Win Rate
4