AI Paper Atlas

Research map / Vision-language and multimodal / Human and embodied modeling

Closing the Loop: Universal Repository Representation with RPG-Encoder

Jane Luo, Chengyu Yin, Xin Zhang, Qingtao Li, Steven Liu, Yiming Huang, Jie Wu, Hao Liu, Yangyu Huang, Yu Kang, Fangkai Yang, Ying Xin and 1 more

ICML 2026 · Vision-language and multimodal › Human and embodied modeling · approach: Layout-aware document understanding · 0 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
add semantic features to the graph (ablation)RPG-Encoder (GPT-4o) · SWE-bench Live function-levelAcc@1: 43.1 -> 50.5Improved
add dependency edges to the graph (ablation)RPG-Encoder (GPT-4o) · SWE-bench Live file-levelAcc@1: 58.4 -> 69.2Improved
add semantic features to the graph (ablation)RPG-Encoder (GPT-4.1) · SWE-bench Live function-levelAcc@1: 57.4 -> 64.7Improved
add dependency edges to the graph (ablation)RPG-Encoder (GPT-4.1) · SWE-bench Live function-levelAcc@1: 63.7 -> 64.7No clear effect
add function metadata to nodes (ablation)ZeroRepo-RPG (GPT-5-mini) · scikit-learn reconstructionPass rate: 74.1 -> 82.8Improved
add file and function metadata to nodes (ablation)ZeroRepo-RPG (GPT-5-mini) · scikit-learn reconstruction#Files: 157 -> 256Improved
add all node metadata (ablation)ZeroRepo-RPG (GPT-5-mini) · scikit-learn reconstructionCoverage: 87.2 -> 100.0Improved
replace API documentation with RPG as reconstruction sourceZeroRepo (GPT-5-mini) · RepoCraftCoverage: 74.2 -> 98.5Improved
replace API documentation with RPG as reconstruction sourceZeroRepo (GPT-4.1) · RepoCraftCoverage: 64.6 -> 93.5Improved
replace baseline localizers with RPG-EncoderClaude-4.5-Sonnet · SWE-bench Verified function-levelAcc@5: 79.3 -> 93.7Improved
replace baseline localizers with RPG-EncoderGPT-5 · SWE-bench Live function-levelAcc@5: 76.2 -> 87.8Improved
use RPG instead of baseline navigation for efficiencyGPT-4.1 · SWE-bench VerifiedAcc@5/Cost: 3.10 -> 4.63Improved

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