Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Scientific Papers

Benchmarks

Task NameDataset NameSOTA ResultTrend
Scientific reasoning graph extraction and introduction generationScientific Papers (test)
Precision/Correlation Score48.6
8
Scientific Peer-ReviewingScientific papers (test)
R0 Score8.7
7
Poster GenerationScientific Papers
Perplexity (PPL)4.6
7
Review Quality EvaluationScientific Papers 200 sampled papers (random sample)
Technical Depth100
6
Scientific ideationscientific papers Out-of-Domain
Win Rate vs GPT-5.259
2
Scientific ideationscientific papers In-Domain
Win Rate vs GPT-5.261
2
Showing 6 of 6 rows