Scientific LLM Benchmarks
GitHub
← All benchmarks
Math· multi-level

MathBench

Shanghai AI Laboratory · 2024

Hierarchical bilingual benchmark spanning arithmetic to college-level theory and application.

Mathematics
GitHub stars
Task type
MCQ
Modality
text
Access
open
Size
3,709 items
License
Apache-2.0
Metrics
Accuracy, CircularEval

Examples

No sample rows available for this dataset.