This repository is the official implementation of the paper LLM-guided Hierarchical Retrieval.
LATTICE proposes an LLM-native retrieval paradigm that combines the efficiency of hierarchical search with the reasoning power of modern large language models. Instead of relying on a static retriever + reranker pipeline or attempting to place a large corpus directly in an LLM context, LATTICE organizes the corpus into a semantic tree and uses an… See the full description on the dataset page:
https://huggingface.co/datasets/quicktensor/lattice-bright-trees.