1 paper · 1 filter
Xiaofeng Lin, Sirou Zhu, Yilei Chen +6
Large language models (LLMs) achieve strong performance when all task-relevant information is available upfront, as in static prediction and instruction-following problems. However…