TAB-PO: Preference Optimization with a Token-Level Adaptive Barrier for Token-Critical Structured Generation

Samah Fodeh, Linhai Ma, Ganesh Puthiaraju, Srivani Talakokkul, Afshan Khan, Ashley Hagaman, Sarah Lowe, Aimee Kendall Roundtree

Published: 2026, Last Modified: 26 May 2026CoRR 2026EveryoneRevisionsBibTeXCC BY-SA 4.0
Loading