AI编程 (5 articles)

Claude 3.5 Sonnet Leads SWE-bench Benchmark, Code Generation Capability Surpasses GPT-4o

Anthropic's Claude 3.5 Sonnet has achieved remarkable performance in the authoritative SWE-bench code benchmark test, successfully surpassing OpenAI's GPT-4o and demonstrating exceptional software engineering capabilities. This breakthrough marks a significant advancement for the Claude series in code generation and provides developers with a more reliable programming assistant.