AI Paper Atlas

Research map / Graph learning / GNN architectures

Retrieval-Augmented Generation for Code Summarization via Hybrid GNN

Shangqing Liu, Yu Chen, Xiaofei Xie, Jing Kai Siow, Yang Liu

ICLR 2021 · Graph learning › GNN architectures · approach: Expressive equivariant message passing · 68 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use retrieval-augmented hybrid graph generation rather than RencosCCSD overall code summarization · CCSD overallBLEU-4: 12.59 -> 14.01Improved
use retrieval-augmented hybrid graph generation rather than RencosCCSD overall code summarization · CCSD overallROUGE-L: 28.45 -> 30.89Improved
use retrieval-augmented hybrid graph generation rather than RencosCCSD overall code summarization · CCSD overallMETEOR: 13.21 -> 14.50Improved
add retrieval augmentation (ablation)HGNN · CCSD overallBLEU-4: 10.26 -> 14.01Improved
add retrieval augmentation (ablation)HGNN · CCSD overallROUGE-L: 27.17 -> 30.89Improved
add static graph (ablation)HGNN · CCSD overallROUGE-L: 30.47 -> 30.89Improved
add dynamic graph (ablation)HGNN · CCSD overallBLEU-4: 13.31 -> 14.01Improved
add both static and dynamic graphs (ablation)HGNN without retrieval augmentation · CCSD overallBLEU-4: 9.98 -> 10.26; 9.93 -> 10.26Improved
use summary-based rather than code-based retrieval augmentation (ablation)HGNN augmentation · CCSD overallBLEU-4: 10.22 -> 13.76Improved
combine code-based and summary-based retrieval augmentation (ablation)HGNN augmentation · CCSD overallBLEU-4: 14.01Improved
use HGNN rather than RencosPCSD code summarization · PCSD testROUGE-L: 36.21 -> 39.91Improved
remove static graph (ablation)HGNN · PCSD testBLEU-4: 24.42 -> 24.06Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy