1 paper · 1 filter
Yuhua Qian
LLM program-search agents are usually judged by the score that drives their search. We study LLM-driven search over Python exercise policies for Bermudan options, an optimal stoppi…