Scientific LLM Benchmarks
GitHub
← All benchmarks
Agentic· ml-research

MLRC-Bench

University of Michigan · 2025

Benchmarks agents on proposing and coding novel methods for seven recent ML research competition problems.

GitHub stars
Task type
agentic
Modality
code
Access
open
Size
7 items
License
MIT
Metrics

Examples

No sample rows available for this dataset.