#artificial intelligence
Sep 2026
Self-Spec Verifiable Code Generation
VeriCodeBench is introduced, a benchmark for self-spec verifiable code generation, where the LLM relies solely on its own generated specification and code throughout the entire process, and CodeNova is introduced to enhance the capabilities of LLMs in self-spec verifiable code generation.
Jia-Ru Qian, Yihong Dong, Yong-Ming Li et al.
· arXiv.org · 0 citations