arXiv:2601.19913v1 Announce Type: cross
Abstract: Distinguishing human-written Korean text from fluent LLM outputs remains difficult even for linguistically trained readers, who can over-trust surface well-formedness. We study whether expert detection can be treated as a learnable skill and improved through structured calibration. We introduce LREAD, a rubric derived from national Korean writing standards and adapted to target micro-level artifacts (e.g., punctuation optionality, spacing behavior, and register shifts). In a three-phase longitudinal blind protocol with Korean linguistics majors, Phase 1 measures intuition-only detection, Phase 2 enforces criterion-level scoring with explicit justifications, and Phase 3 evaluates domain-focused mastery on held-out elementary essays. Across phases, majority-vote accuracy increases from 60% to 100%, accompanied by stronger inter-annotator agreement (Fleiss’ kappa: -0.09 –> 0.82). Compared to state-of-the-art LLM detectors, calibrated humans rely more on language-specific micro-diagnostics that are not well captured by coarse discourse priors. Our findings suggest that rubric-scaffolded expert judgment can serve as an interpretable complement to automated detectors for non-English settings, and we release the full rubric and a taxonomy of calibrated detection signatures.
