作者位次 #5/8ACL 2026Poster· Natural Language Processing
C ode F low B ench: A Multi-turn, Iterative Benchmark for Complex Code Generation
Sizhe Wang · Zhengren Wang · Dongsheng Ma · Yongan Yu · Rui Ling · Zhiyu li · Feiyu Xiong · Wentao Zhang
摘要
Modern software development demands code that is maintainable, testable, and scalable by organizing the implementation into modular components with iterative reuse of existing codes. We formalize this iterative, multi-turn paradigm as codeflow and introduce CodeFlowBench , the first benchmark des...