I built an LLM benchmark harness that lets you browse and compare how models answered each question

Hello everyone. I felt current LLM benchmark harnesses hand you headline numbers but offer no tooling to see how models actually answered each question (they dump everything to a JSONL or Parquet file, so you end up writing custom code just to read the answers). So I built lm-eval-ledger: a be

中文全文翻译正在补齐，可先查看下方英文原文。

Hello everyone. I felt current LLM benchmark harnesses hand you headline numbers but offer no tooling to see how models actually answered each question (they dump everything to a JSONL or Parquet file, so you end up writing custom code just to read the answers). So I built lm-eval-ledger: a be