1 paper · 1 filter
Arpan Mukherjee, Marcello Bullo, Debabrota Basu +1
While test-time scaling with verification has shown promise in improving the performance of large language models (LLMs), the role of the verifier and its imperfections remain unde…