1 paper · 1 filter
Charles Schepanowski, Charles Ling
Large language models (LLMs) achieve impressive results on many benchmarks, yet their capacity for planning and stateful reasoning remains unclear. We study these abilities directl…