1 paper
Avijit Ghosh, Anka Reuel, Jenny Chim +45
AI evaluation results are produced at scale but reported inconsistently across leaderboards, model cards, benchmark papers, and company blogs. The cost is interpretive: readers can…