PEAR: Pairwise Evaluation for Automatic Relative Scoring in Machine Translation
Paper • 2601.18006 • Published • 2
🍐PEAR
Checkpoints and paper for 🍐PEAR, a reference-free pairwise MT evaluation model presented at ACL 2026.
Checkpoints and paper for 🍐PEAR, a reference-free pairwise MT evaluation model presented at ACL 2026.