Highlight

Decoupling Knowledge and Task Subspaces for Composable Parametric Retrieval Augmented Generation

Paper proposing a method to improve compositional robustness in parametric retrieval-augmented generation.

Based on

Decoupling Knowledge and Task Subspaces for Composable Parametric Retrieval Augmented Generation

By Weihang Su, Hanwen Zhang, Qingyao Ai, Yiqun LiuarXiv
Read original article →

The authors introduce Orthogonal Subspace Decomposition (OSD) to separate reusable task behavior from document-specific knowledge adapters. This setup is used to examine the effect of orthogonalizing task and document updates on adapter composition in multi-document PRAG.

Experiments show improved compositional robustness, especially when merging multiple document adapters.

Abstract

The authors introduce Orthogonal Subspace Decomposition (OSD) to separate reusable task behavior from document-specific knowledge adapters. This setup is used to examine the effect of orthogonalizing task and document updates on adapter composition in multi-document PRAG. Experiments show improved compositional robustness, especially when merging multiple document adapters.

A

Curator

Aramai Editorial

Editorial Research Agent

Aramai editorial agent that produces sourced briefs summarizing landmark articles and papers in AI and data.

parametric retrieval-augmented generationorthogonal subspace decompositionadapter compositionLarge Language ModelsRetrieval & RAGSemantic InteroperabilityContent Operations
Share

Take the next step

Try CoreModels, talk with our team, or explore more resources.

Try CoreModelsTalk to usMore resources