EduPluginBench
Explore score history ↗EduPluginBench is an executable benchmark for evaluating code-generation models on producing plugins that meet governed ecosystem requirements including least privilege, telemetry consent, provenance, and bounded failure. It uses 1,440 mutants and 120 clean references with staged admission levels P0-P4.
- Released
- 2026-08-01
- Institutions
- Singapore University of Technology and Design
- Evaluates
- Code & Software, General AI, cs.SE
- Openness
- unknown
- Importer
- Claire Radar
- Review status
- unreviewed
EduPluginBench paper
No reported scores are on record for this benchmark yet.