MultiLexNorm++ (LLM Pipeline) - Korean: leaderboard

Metric: Error reduction rate (%; (correct minus wrong normalizations) / words that need normalization, so 0 equals leaving the text unchanged and a negative score means more wrong than right edits; the floor is -100 divided by the share of words needing normalization; 8-shot prompting, mean of three runs; the LLM normalizes the words an XLM-R detector flags after a training-lexicon lookup; the new Korean social-media set, 7.52% of words normalized). Source: arxiv.org. Saturation forecast: Estimated already saturated. 4 models tracked.

Top models

#ModelScore
1Qwen 2.5 72B Instruct-6.88
2Gemma 3 27B (IT)-8.73
3GPT-4o-8.73
4Llama 3.3 70B Instruct-14.82

Interactive version: theaggregate.ai/benchmark?slug=multilexnorm-plus-plus-llm-pipeline-korean · How It Works · Data refreshed daily, snapshot 2026-09-26.