作者位次 #5/8ACL 2026Findings· Findings of ACL
Lost in Translation: Do LVLM Judges Generalize Across Languages?
Md Tahmid Rahman Laskar · Mohammed Saidul Islam · Mir Tafseer Nayeem · Amran Bhuiyan · Mizanur Rahman · Shafiq Joty · Enamul Hoque · Jimmy Xiangji Huang
摘要
Automatic evaluators such as reward models play a central role in the alignment and evaluation of large vision–language models (LVLMs). Despite their growing importance, these evaluators are almost exclusively assessed on English-centric benchmarks, leaving open the question of how well these eva...