{"id":"2404.14367","title":"Preference Fine-Tuning of LLMs Should Leverage Suboptimal, On-Policy Data","url":"https://codeofpaper.com/paper/2404.14367","arxiv_url":"https://arxiv.org/abs/2404.14367","published":"2024-04-22","code":"community","best":{"full_name":"Asap7772/understanding-rlhf","url":"https://github.com/Asap7772/understanding-rlhf","tier":"high_confidence_community","stars":32,"license":"MIT","framework":null,"language":"Python","last_commit":"2024-04-20T18:42:50"},"most_starred_community":null,"implementations":[{"full_name":"Asap7772/understanding-rlhf","url":"https://github.com/Asap7772/understanding-rlhf","tier":"high_confidence_community","stars":32,"license":"MIT","framework":null,"language":"Python","last_commit":"2024-04-20T18:42:50"}],"possible_matches":1}