K-HALU: Multiple Answer Korean Hallucination Benchmark for Large Language Models

Jaehyung Seo, Heu-Jeoung Lim
2/12/2026

No official code found. Best implementation: J-Seo/K-HALU (38 stars, MIT, last commit 2025-12-30).

Semantic Scholar

Code Implementations(1)

J-Seo/K-HALUCommunity65%
MIT

K-HALU: Multiple Answer Korean Hallucination Benchmark for Large Language Models

3812C++, ShellFeb 21, 20259 months agoMIT

Cite this paper

@article{seo2026khalu,
  title  = {K-HALU: Multiple Answer Korean Hallucination Benchmark for Large Language Models},
  author = {Jaehyung Seo and Heu-Jeoung Lim},
  year   = {2026},
  url    = {https://api.semanticscholar.org/CorpusID:b6692fbc2cf9eed9ebcd4d392f4d2c436f07f6d7},
  journal = {ICLR 2025 2025}
}

Discussion