AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Seeking Commonality, Preserving Specificity: A Spectral-Aware Hierarchical Framework for Cross-City Road Representation Learning

Jingtian Ma, Jingyuan Wang, Leong Hou U

ICML 2026 · Label-efficient and robust learning › Recognition and long-tail · approach: Correspondence and causal representation learning

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 region embedding (hierarchical region abstraction) (ablation)CoSpec · XA, all three tasksaccuracy / F1Improved
add city embedding (ablation)CoSpec · XA next location predictionACC@1Improved
add low-frequency path (ablation)CoSpec · XA label classification and destination predictionF1 / ACC@1Improved
add high-frequency path (ablation)CoSpec · XA next location predictionACC@1Improved
add orthogonality loss (ablation)CoSpec · XAall metricsImproved
add reconstruction loss (ablation)CoSpec · XAall metricsImproved
add graph smoothness loss (ablation)CoSpec · XAall metricsImproved
add prototype alignment loss (ablation)CoSpec · XAall metricsImproved
set number of regions N_R to 200 (ablation)CoSpec · XA sensitivityACC@1 / F1Improved
set number of prototypes N_P to 100 (ablation)CoSpec · XA sensitivityACC@1 / F1Improved
increase low-frequency path order K_L up to 3 or 4 (ablation)CoSpec · XA sensitivityACC@1 / F1Improved
increase high-frequency path order K_H beyond 2 (ablation)CoSpec · XA sensitivityACC@1 / F1Hurt

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