List your company

Repo-level Compilation on CompileAgentBench

96CSRCompileAgent
22.1641.3360.579.67May 7, 2025

Evaluation results

MethodLinks
CompileAgentBackbone=Claude-3-5-sonnet, Size=-2025.05965.3722.02
CompileAgentBackbone=DeepSeek-v2.5, Size=236B2025.059111.383.31
CompileAgentBackbone=GPT-4o, Size=-2025.05898.3816.53
CompileAgentBackbone=Qwen2.5-32B-Instruct, Size=32B2025.05805.253.16
Readme-AIBackbone=Claude-3-5-sonnet, Size=-2025.0579127.3355.26
CompileAgentBackbone=LLama3.1-70B, Size=70B2025.05797.382.71
RAGBackbone=Claude-3-5-sonnet, Size=-2025.05788.354.44
Readme-AIBackbone=GPT-4o, Size=-2025.0572128.842.94
RAGBackbone=DeepSeek-v2.5, Size=236B2025.057211.336.08
Readme-AIBackbone=DeepSeek-v2.5, Size=236B2025.0571125.4333.7
Readme-AIBackbone=Qwen2.5-32B-Instruct, Size=32B2025.0570127.8233.18
RAGBackbone=GPT-4o, Size=-2025.056711.1245.78
CompileAgentBackbone=Gemini-1.5-flash, Size=-2025.05653.552.39
RAGBackbone=Qwen2.5-32B-Instruct, Size=32B2025.056210.5536.73
Readme-AIBackbone=LLama3.1-70B, Size=70B2025.0561125.0333.57
RAGBackbone=LLama3.1-70B, Size=70B2025.056110.9836.87
CompileAgentBackbone=Mixtral-8×7B-Instruct, Size=42B2025.05554.884.32
RAGBackbone=Gemini-1.5-flash, Size=-2025.05469.2835.72
RAGBackbone=Mixtral-8×7B-Instruct, Size=42B2025.054510.8236.49
Readme-AIBackbone=Gemini-1.5-flash, Size=-2025.0541123.6832.37
Readme-AIBackbone=Mixtral-8×7B-Instruct, Size=42B2025.0538124.633.12
Oss-Fuzz-GenNote=Operates without relying on LLMs2025.052553.01—