1 paper · 1 filter
Zhouhang Xie, Tushar Khot, Bhavana Dalvi Mishra +4
Instruction-following LLMs have recently allowed systems to discover hidden concepts from a collection of unstructured documents based on a natural language description of the purp…