BWhat changed
Integrated a balanced Korean/multilingual/code pretraining mixture with Korean-centered human demonstration/preference alignment, documenting bilingual and cross-lingual evaluation in one model family.
Scientific scope
Integrated a balanced Korean/multilingual/code pretraining mixture with Korean-centered human demonstration/preference alignment, documenting bilingual and cross-lingual evaluation in one model family. The work builds on HyperCLOVA and established decoder, positional-encoding, grouped-query and preference-training methods. Its contribution is the Korean-centered training recipe and reusable model resource; the evaluations do not guarantee cultural correctness or safety.
Evidence
The linked original sources describe the stated contribution and its evidence limits. WRS has not independently rerun the experiments or certified a proof. The work builds on HyperCLOVA and established decoder, positional-encoding, grouped-query and preference-training methods. Its contribution is the Korean-centered training recipe and reusable model resource; the evaluations do not guarantee cultural correctness or safety.
Contributors
- Source-listed authors and technical contributors · SCIENCE: Source-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities.
Provisional WRS assessment, edition 2026-10-16.1. Independent expert calibration is pending. This grade does not certify correctness or assign ownership.
Full scientific record · Browse contribution grades · Request a correction