AI Paper Atlas

Research map / Graph learning / Graph algorithms and clustering

Skill Path: Unveiling Language Skills from Circuit Graphs

Hang Chen, Xinyu Yang, Jiaying Zhu, Wenya Wang

AAAI 2026 · Graph learning › Graph algorithms and clustering · approach: Graph distance and retrieval · 4 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
remove skill paths of the target skill from the circuit graph (ablation)GPT2-small circuit graph · PVT/IDT/ICL samplesaccuracy vs original output: PVT 1.00 -> 0.01; IDT 1.00 -> 0.00; ICL1 1.00 -> 0.00Hurt
remove ~500 random paths instead of skill paths (ablation)GPT2-small circuit graph · PVT/IDT/ICL samplesaccuracy vs original output: PVT 1.00 -> 0.23 (random 500); IDT 1.00 -> 0.29Mixed
combine existing pruning methods with skill-path extractionACDC/EAP/DiscoGP/etc. · skill overlapovlp(IDT, PVT): ACDC 0.19 -> 0.74; DiscoGP 0.18 -> 0.79Improved
combine pruning with skill-path extractionACDC · skill overlapovlp(ICL1, IDT): 0.17 -> 0.63Improved
use skill paths for unrelated greater-than skillskill paths · GT vs PVT/IDT/ICL1edge overlap: ovlp(GT,PVT) ACDC 0.11 -> 0.68; ovlp(GT,IDT) 0.05 -> 0.13Improved
use ACDC, DiscoGP, Scrubbing instead of E-pruning, EAP as pruning (ablation)skill path framework · layer distribution of receiversstratificationMixed

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