Language Model Evaluation Harness Open-source framework that evaluates language models across academic benchmarks, supporting reproducible comparisons with shared tasks, prompts, and metrics.