AI RESEARCH

Fluent Alignment with Disfluent Judges: Post-training for Lower-resource Languages

arXiv CS.AI

ArXi:2512.08777v2 Announce Type: replace-cross We propose a post-