TY  - RPRT
TI  - LLM2IR: simple unsupervised contrastive learning makes long-context LLM great retriever
AU  - Xiaocong Yang
PY  - 2025
UR  - https://arxiv.org/abs/2601.05262
ID  - 2601.05262
ER  -