Atrex-Bench
Explore score history ↗Evaluates coding agents on generating GPU kernels from PyTorch references across 30 operators and 440 shapes derived from production inference traces. Scoring uses a three-stage evaluator measuring compile success, numerical correctness, and speed-of-light (SOL) efficiency against a cached roofline. The benchmark includes 4 DSL backends and supports multi-vendor GPUs.
- Released
- 2026-07-16
- Institutions
- Alibaba Group (China)
- Evaluates
- Agents, Language & Knowledge, Software & AI Compute
- Openness
- unknown
- Importer
- Claire Radar
- Review status
- unreviewed
Atrex-Bench paper and code
No reported scores are on record for this benchmark yet.