Unverified50% confidenceBenchmarkExact time
2021 年 Codex 论文（Chen et al.）在 GitHub 代码语料上完成专项预训练，并提出了 HumanEval 基准测试
1
Sources
50%
Confidence
Long-term
Relevance
7/21/2026
First Seen
Sources
Related Claims
UnverifiedHumanEval基准测试由OpenAI于2021年随Codex模型论文一同发布，包含164道编程题，Pass@1指一次生成即通过全部测试用例的比例67% similarUnverifiedCodex模型的评估通常使用HumanEval和MBPP等编码基准测试66% similarVerifiedThe training data sources for specialized programming models include open-source code from GitHub, competitive programming solutions, technical documentation, and code review records, and the models are optimized for benchmarks such as HumanEval, MBPP, and SWE-bench.66% similarUnverifiedHumanEval was developed in-house by OpenAI and is specifically designed to test the functional correctness rate of code generation.65% similarUnverifiedThe integration of GitHub with Claude Code can automatically analyze the scope of changes, identify potential risk points, and suggest test coverage when a PR is submitted63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/584716API
curl https://kongchang.com/api/v1/knowledge/claims/584716MCP
get_claim(id=584716)