Kernel Bench L3

coding

Multilingual: no

Kernel Bench L3 evaluates agentic GPU kernel optimization across 50 problems. Qwen reports two metrics for this benchmark: median per-problem speedup over the PyTorch eager reference and the fraction of problems faster than torch.compile.

Methodology

Imported from llm-stats public benchmark metadata. Modality: text. Max score: 1. Categories: agents, coding, systems. Language: en. Verified by llm-stats: no.