Three tasks evaluation scores across different disciplines
Reference Generation: Precision
Abstract Writing: True
Review Composition: Precision
Literature Review Writing: KPR
Radar chart of the accuracy of LLM-generated references across various dimensions
Compare with Semantic Scholar(T1)
Compare with Semantic Scholar(T3)
Compare with Human-cited References(T1)
Compare with Human-cited References(T3)
