Unverified50% confidenceFactExact time
Compared to benchmarks like HumanEval that only examine single-function code generation, SWE-bench places greater emphasis on cross-file understanding and multi-step planning.
1
Sources
50%
Confidence
Long-term
Relevance
9/12/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedSWE-RPG在SWE-bench基础上引入更细粒度的评分维度，对需求理解、规划质量、代码风格等中间环节进行独立打分71% similarUnverified大语言模型编程场景的核心能力指标包括HumanEval代码生成基准、SWE-bench真实Bug修复成功率及跨文件依赖解析准确性68% similarUnverifiedSWE-bench比HumanEval（单函数补全）或MBPP（简单算法题）更贴近生产环境68% similarUnverified结构化的提示词相比纯文本段落，能让模型更精确地识别不同指令模块的边界，减少指令混淆68% similarUnverified简单代码补全用轻量模型即可，涉及跨文件重构、架构级推理时才需要旗舰能力67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/915328API
curl https://kongchang.com/api/v1/knowledge/claims/915328MCP
get_claim(id=915328)