Tom Sherborne

A comprehensive benchmarking tool to assess AI model efficiency in real-world scenarios.

Examining the impact of data contamination on code generation evaluations.

Transform discarded models into powerful new solutions through model merging.