AI Paper Atlas

Research map / Large language models / Retrieval and knowledge

MobileKGQA: On-Device KGQA System on Dynamic Mobile Environments

Junyong Ahn, Hyeongrok Han, Bong Gyun Kang, Jisoo Mok, Byunghan Lee, Sungroh Yoon

ICLR 2026 · Large language models › Retrieval and knowledge · approach: Knowledge graph reasoning with LLMs

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
compress embeddings to 256-bit hash codes (ablation)MobileKGQA · WebQSPHit: 79.5 -> 79.5No clear effect
compress embeddings to 64-bit hash codes (ablation)MobileKGQA · CWQHit: 54.4 -> 53.5Hurt
use 256-bit hash codes instead of float embeddings for storage (ablation)MobileKGQA · WebQSP embedding storageDisk (GB): 6.2 -> 0.065Improved
add stepwise sequential-reasoning annotation generation instead of CoT (ablation)MobileKGQA adaptation · WebQSP D1 to D2Hit total: 61.4 -> 64.9Improved
add stepwise annotation generation instead of RLM (ablation)MobileKGQA adaptation · WebQSP D1 to D2Hit total: 62.3 -> 64.9Improved
use stepwise annotation generation instead of CoT for token cost (ablation)annotation generation · WebQSP (Gemma 2 2B)output tokens: 328687 -> 97855Improved
use hash-code MobileKGQA instead of GNN-RAG on Jetson 7WKGQA training · WebQSP Jetson 7Wtraining energy (Wh): 28.7 -> 11.8Improved
use hash-code MobileKGQA instead of tuned-down GNN-RAG on JetsonKGQA training · WebQSP Jetson 7WHit: 61.7 -> 74.2Improved
add annotation-based adaptation to new domain D3 dataMobileKGQA · CWQ (S1 to S1(PA))Hit: 57.8 -> 60.2Improved
use MobileKGQA instead of GNN-RAG under 80% triple deletionKGQA · WebQSP 80% deletionHit: 56.7 -> 71.4Improved
use MobileKGQA annotation instead of symKGQA few-shotKGQA adaptation · WebQSP zero-shot (Gemma 2 2B)Hit: 10.6 -> 37.3Improved
use MobileKGQA instead of GNN-RAG (GTE-large)KGQA · WebQSP (Gemma 2 2B)Hit: 80.0 -> 79.8No clear effect

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