Review
Jul 2026
Do You Need a Frontier Model as a Citation Verifier? Benchmarking Rubric LLMs for Deep-Research Source Attribution
Calibrating the judge is a prerequisite for using citation rubrics as reward signals, and the results show that this calibration does not require the most expensive available model.
E. Leung, Elias Lumer, Corey Feld et al.
· arXiv.org · 0 citations