MindRef: Mimicking Human Memory for Hierarchical Reference Retrieval with Fine-Grained Location Awareness
This addresses the need for efficient reference retrieval in knowledge-intensive tasks, but it is incremental as it builds on existing LLM capabilities with constrained decoding.
The paper tackles the problem of retrieving reference passages from large language models without needing pre-segmented chunks, by proposing a two-stage framework that mimics human memory to recall document titles and fine-grained passages, achieving verified improvements in downstream tasks on KILT benchmarks.
When completing knowledge-intensive tasks, humans sometimes need an answer and a corresponding reference passage for auxiliary reading. Previous methods required obtaining pre-segmented article chunks through additional retrieval models. This paper explores leveraging the parameterized knowledge stored during the pre-training phase of large language models (LLMs) to recall reference passage from any starting position independently. We propose a two-stage framework that simulates the scenario of humans recalling easily forgotten references. Initially, the LLM is prompted to recall document title identifiers to obtain a coarse-grained document set. Then, based on the acquired coarse-grained document set, it recalls fine-grained passage. In the two-stage recall process, we use constrained decoding to ensure that content outside of the stored documents is not generated. To increase speed, we only recall a short prefix in the second stage, and then locate its position to retrieve a complete passage. Experiments on KILT knowledge-sensitive tasks have verified that LLMs can independently recall reference passage locations in various task forms, and the obtained reference significantly assists downstream tasks.