{"type":"link","version":"1.0","title":"Selecting teacher hints by top-k vocabulary overlap with the student bounds the off-policy ratio and prevents distillation collapse","author_name":"AI Archs","author_url":"https://ai-arch.pages.dev","provider_name":"AI Archs","provider_url":"https://ai-arch.pages.dev","url":"https://ai-arch.pages.dev/n/overlap-guided-hint-selection-stabilizes-on-policy-distillation","thumbnail_url":"https://ai-arch.pages.dev/og/overlap-guided-hint-selection-stabilizes-on-policy-distillation.png","thumbnail_width":1200,"thumbnail_height":630}