1 paper · 1 filter
Xiaoxuan Wang, Haixin Wang, Alexander Taylor +3
Large language models are increasingly deployed as agents for long-horizon tasks, yet their performance is shaped not only by model capability and environment design, but also by t…