Current Large Audio Language Models largely transcribe rather than listenarxiv.org4 points·earcar··0 commentsOpen articleSaveView on HN