SciCode: A Research Coding Benchmark Curated by Scientists

Since language models (LMs) now outperform average humans on many challenging tasks, it has become increasingly difficult to develop challenging, high-quality, and realistic evaluations. We address this issue by examining LMs' capabilities to generate code for solving real scientific research p...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:arXiv.org 2024-07
Hauptverfasser: Tian, Minyang, Gao, Luyu, Zhang, Shizhuo Dylan, Chen, Xinan, Fan, Cunwei, Guo, Xuefei, Haas, Roland, Pan, Ji, Krongchon, Kittithat, Yao, Li, Liu, Shengyan, Luo, Di, Ma, Yutao, Tong, Hao, Trinh, Kha, Tian, Chenyu, Wang, Zihan, Wu, Bohao, Xiong, Yanyu, Yin, Shengzhu, Zhu, Minhui, Kilian Lieret, Lu, Yanxin, Liu, Genglin, Du, Yufeng, Tao, Tianhua, Press, Ofir, Callan, Jamie, Huerta, Eliu, Peng, Hao
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Since language models (LMs) now outperform average humans on many challenging tasks, it has become increasingly difficult to develop challenging, high-quality, and realistic evaluations. We address this issue by examining LMs' capabilities to generate code for solving real scientific research problems. Incorporating input from scientists and AI researchers in 16 diverse natural science sub-fields, including mathematics, physics, chemistry, biology, and materials science, we created a scientist-curated coding benchmark, SciCode. The problems in SciCode naturally factorize into multiple subproblems, each involving knowledge recall, reasoning, and code synthesis. In total, SciCode contains 338 subproblems decomposed from 80 challenging main problems. It offers optional descriptions specifying useful scientific background information and scientist-annotated gold-standard solutions and test cases for evaluation. Claude3.5-Sonnet, the best-performing model among those tested, can solve only 4.6% of the problems in the most realistic setting. We believe that SciCode demonstrates both contemporary LMs' progress towards becoming helpful scientific assistants and sheds light on the development and evaluation of scientific AI in the future.
ISSN:2331-8422