Smoothie-Qwen reduces Chinese-language output in Qwen2.5-Coder by scaling down lm head weights of Chinese tokens, reaching 95% suppression on a synthetic elicitation set while keeping Korean MMLU accuracy roughly stable.
Bias and fairness in large language models: A survey
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Smoothie-Qwen: Post-Hoc Smoothing to Reduce Language Bias in Multilingual LLMs
Smoothie-Qwen reduces Chinese-language output in Qwen2.5-Coder by scaling down lm head weights of Chinese tokens, reaching 95% suppression on a synthetic elicitation set while keeping Korean MMLU accuracy roughly stable.