Search papers, labs, and topics across Lattice.
University of Waterloo
2
0
6
2
LLM-generated text alone can be a surprisingly effective and cost-efficient source of feedback for pseudo-relevance feedback, rivaling corpus-derived feedback in low-resource information retrieval tasks.
NanoKnow reveals that even with external evidence, LLMs are more accurate when answers were seen during pre-training, highlighting the crucial role of parametric knowledge.
