TY  - RPRT
TI  - metabench -- A Sparse Benchmark of Reasoning and Knowledge in Large Language Models
AU  - Alex Kipnis
AU  - Konstantinos Voudouris
AU  - Luca M. Schulze Buschoff
AU  - Eric Schulz
PY  - 2025
UR  - https://arxiv.org/abs/2407.12844
ID  - 2407.12844
ER  -