1 paper · 1 filter
Paul Landes, Pranav Herur, Adam Cross +1
Large Language Models (LLMs) are optimized to produce distributionally plausible continuations rather than to explicitly verify whether generated propositions are entailed by sourc…