The Open ASR Leaderboard now filters out "benchmaxxing" submissions. This move aims to improve the reliability and fairness of speech recognition model evaluations. Developers can trust the leaderboard to showcase genuine performance gains.
Opening Kapyn…