{"type":"link","version":"1.0","title":"Preference Fine-Tuning of LLMs Should Leverage Suboptimal, On-Policy Data","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/arxiv2404.14367","thumbnail_url":"https://ai-arch.pages.dev/og/arxiv2404.14367.png","thumbnail_width":1200,"thumbnail_height":630}