AI Paper Atlas

Research map / Large language models / Retrieval and knowledge

HanjaBridge: Resolving Semantic Ambiguity in Korean LLMs via Hanja-Augmented Pre-Training

Seungho Choi, Sihyun Park, Minsang Kim, Chansol Park, Bongsu Kim

AAAI 2026 · Large language models › Retrieval and knowledge · approach: Knowledge editing in 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
add Hanja candidates (k=8) on top of distillation-only CPT (ablation)Qwen2.5-3B CPT · KoBALT-Hardavg accuracy: 0.1028 -> 0.1437Improved
add Hanja candidates (k=8) on top of distillation-only CPT (ablation)Qwen2.5-3B CPT · KoBEST-Generalavg accuracy: 0.5690 -> 0.6354Improved
add token-level distillation to CPT (ablation)Qwen2.5-3B CPT · English averageavg accuracy: 0.6501 -> 0.6753Improved
add Hanja candidates (k=8) to distillation CPT (ablation)Qwen2.5-3B CPT · English averageavg accuracy: 0.6753 -> 0.6792Improved
increase number of Hanja candidates from 8 to 16 (ablation)HanjaBridge · KoBALT-Hardavg accuracy: 0.1437 -> 0.1322Hurt
increase number of Hanja candidates from 2 to 8 (ablation)HanjaBridge · KoBALT-Hardavg accuracy: 0.0983 -> 0.1437Improved
plain full CPT on Korean data without distillation or HanjaQwen2.5-3B · KoBALT-Hardavg accuracy: 0.1189 -> 0.1115Hurt
plain full CPT on Korean data without distillation or HanjaQwen2.5-3B · English averageavg accuracy: 0.6976 -> 0.6501Hurt
use Hanja candidates at inference in addition to training (ablation)HanjaBridge · Hanja multiple-choice probeaccuracy: 0.568 -> 0.575No clear effect
train with HanjaBridge, Korean-only at inference, versus Full CPT (ablation)Qwen2.5-3B · Hanja multiple-choice probeaccuracy: 0.389 -> 0.568Improved
longer training with Hanja augmentationHanjaBridge k=8 · Hanja attention probecorrect-candidate attention accuracy: 0.102 -> 0.308Improved

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