Show HN: We Westernized Qwen for US enterprise use (huggingface.co)

Hirundo has announced the release of a standalone merged checkpoint based on the Qwen/Qwen3.6-35B-A3B model, specifically tailored for US enterprise usage. This model, built upon a behavioral-unlearning approach, significantly reduces censorship and geopolitical biases while maintaining desired behaviors. The results show a remarkable decrease in bias-related issues, with censorship and propaganda reduction from 89.8% in the base model to just 2.8% in Hirundo's version, alongside substantial declines in refusal and non-compliance rates. This development is significant for the AI/ML community as it addresses critical concerns regarding AI accountability and ethical behavior, particularly in products destined for sensitive markets. Technical implications include the model's capability to operate without needing separate LoRA/PEFT adapters, streamlining the deployment process. Moreover, the detailed evaluation against the new CCPC-500 benchmarks highlights Hirundo's commitment to creating AI systems that prioritize safety and compliance, which could influence the development of future AI models by setting new standards in bias reduction and ethical AI use.