Benchmark
Atrex-Bench
What is Atrex-Bench?
Evaluates coding agents on generating GPU kernels from PyTorch references across 30 operators and 440 shapes derived from production inference traces. Scoring…
- Released
- 2026-07-16
- Evaluates
- Agents, Language & Knowledge, Software & AI Compute
- Openness
- unknown
- Importer
- Claire Radar
- Review status
- unreviewed
- Reported scores
- 0
Source provenance
- Original evidence https://arxiv.org/abs/2607.14541
Atrex-Bench paper and code
No reported scores are on record for this benchmark yet.