Distilling DeepSeek into GPT-OSS skips censorship, scores 83.61%

Using DeepSeek V4 Flash as teacher at an 8k token budget, the self-distilled GPT-OSS-120B scored 83.61% on FinanceReasoning, beating Kimi K3 (81.93%) and Inkling (65.13%). The team released the 20B open weights.