GRAFT (Gated Retrieval Activation Framework for Trees) is an adaptive, complexity-aware Retrieval-Augmented Generation system that grafts specialist reasoning modules onto a RAPTOR-style document tree, activating only the modules and retrieval depth each query actually needs.

2 Open Issues Need Help Last updated: Aug 14, 2026

Open Issues Need Help

View All on GitHub
good first issue indexing

GRAFT (Gated Retrieval Activation Framework for Trees) is an adaptive, complexity-aware Retrieval-Augmented Generation system that grafts specialist reasoning modules onto a RAPTOR-style document tree, activating only the modules and retrieval depth each query actually needs.

Python
documentation good first issue

GRAFT (Gated Retrieval Activation Framework for Trees) is an adaptive, complexity-aware Retrieval-Augmented Generation system that grafts specialist reasoning modules onto a RAPTOR-style document tree, activating only the modules and retrieval depth each query actually needs.

Python