一作ICLR 2026Poster
High Accuracy, Less Talk (HALT): Reliable LLMs through Capability-Aligned Finetuning
Tim Franzmeyer · Archie Sravankumar · Lijuan Liu · Yuning Mao · Rui Hou · Sinong Wang · Jakob Foerster · Luke Zettlemoyer · Madian Khabsa
对齐与偏好OpenReview