{"about":{"non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","site":"https://codewithpapers.app","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page","syntology":{"site":"https://syntology.ai","developers":"https://syntology.ai/developers","mcp":{"server":"https://syntology.ai/mcp","transport":"streamable-http","server_card":"https://syntology.ai/.well-known/mcp/server-card.json","auth":{"type":"trial token, no account","trial_token":"https://syntology.ai/api/oauth/trial/token","method":"POST","docs":"https://syntology.ai/developers"}},"have":"https://syntology.ai/api/graph/have?x=<method, arXiv id or title> (free, answers coverage only)","paper_base":"https://syntology.ai/paper/","atlas_base":"https://app.syntology.ai/?focus="},"machine_readable":[{"url":"https://codewithpapers.app/llms.txt","what":"the machine catalog: every machine-readable file, counted"},{"url":"https://codewithpapers.app/index/manifest.json","what":"paper-to-code index by arXiv id, with Syntology's counts"},{"url":"https://codewithpapers.app/search/manifest.json","what":"site search index (titles, authors) and its files"},{"url":"https://codewithpapers.app/download","what":"bulk files: Syntology's layer, described there"},{"url":"https://codewithpapers.app/build_manifest.json","what":"the build record: inputs, counts, exclusions, probes"}]},"url":"/paper/fine-tuning-language-models-with-reward","title":"Fine-Tuning Language Models with Reward Learning on Policy","arxiv_id":"2403.19279","date":"2024-03-28","proceeding":null,"authors":["Hao Lang","Fei Huang","Yongbin Li"],"abstract":"Reinforcement learning from human feedback (RLHF) has emerged as an effective approach to aligning large language models (LLMs) to human preferences. RLHF contains three steps, i.e., human preference collecting, reward learning, and policy optimization, which are usually performed serially. Despite its popularity, however, (fixed) reward models may suffer from inaccurate off-distribution, since policy optimization continuously shifts LLMs' data distribution. Repeatedly collecting new preference data from the latest LLMs may alleviate this issue, which unfortunately makes the resulting system more complicated and difficult to optimize. In this paper, we propose reward learning on policy (RLP), an unsupervised framework that refines a reward model using policy samples to keep it on-distribution. Specifically, an unsupervised multi-view learning method is introduced to learn robust representations of policy samples. Meanwhile, a synthetic preference generation approach is developed to simulate high-quality preference data with policy outputs. Extensive experiments on three benchmark datasets show that RLP consistently outperforms the state-of-the-art. Our code is available at \\url{https://github.com/AlibabaResearch/DAMO-ConvAI/tree/main/rlp}.","url_abs":"https://arxiv.org/abs/2403.19279v1","url_pdf":"https://arxiv.org/pdf/2403.19279v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"fine-tuning-language-models-with-reward","repo_url":"https://github.com/alibabaresearch/damo-convai","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"multi-view-learning","task_name":"MULTI-VIEW LEARNING"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/2403.19279","atlas_url":"https://app.syntology.ai/?focus=2403.19279","mcp":null,"developers":"https://syntology.ai/developers","held":true,"reason":"arxiv_id_in_atlas","checked_against":{"atlas_ids":549449,"atlas_ids_sha256":"afd72399dafd9906e9062927f70655952385a3d75488558e1f0484e950ab09b7","read_at":"2026-10-05T01:46:21Z"},"agent_calls":[{"tool":"get_citation_path","arguments":{"paper_1":"2403.19279"},"arguments_in":{"paper_2":"another paper's arXiv id or title"},"call":"get_citation_path(paper_1=\"2403.19279\", paper_2=\"…\")"},{"tool":"get_concepts_for_paper","arguments":{"arxiv_id":"2403.19279"},"call":"get_concepts_for_paper(arxiv_id=\"2403.19279\")"}],"by_repo":[{"repo":"alibabaresearch/damo-convai","kind":"official","samples":0,"ran":0,"instrument":0,"unverified":0,"harvested_for_other_papers":3,"harvested_for":["2210.15332","2305.14839","2501.01821"],"harvested_for_calls":[{"arxiv_id":"2210.15332","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2210.15332"},"call":"get_harvested_code_for_paper(arxiv_id=\"2210.15332\")"},{"arxiv_id":"2305.14839","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2305.14839"},"call":"get_harvested_code_for_paper(arxiv_id=\"2305.14839\")"},{"arxiv_id":"2501.01821","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2501.01821"},"call":"get_harvested_code_for_paper(arxiv_id=\"2501.01821\")"}]}],"by_repo_harvested_for_is":"harvested_for: the arXiv ids (as Syntology's graph holds them) of every paper whose samples Syntology's graph links to this repository, in arXiv-id order, complete; the page names the first 3 and says how many more. A link only: it does not say the papers are the same work, that one copied the other, or that the code implements this paper's method; harvested_for_calls gives the get_harvested_code_for_paper call for each arXiv-shaped id among them","by_repo_is":"one row per repository in the archive's code links, Syntology's graph links or the samples; samples 0 means no sample is linked to this paper, a statement about Syntology's coverage, not about the repository: with harvested_for_other_papers present, Syntology harvested the repository for that many other papers, and without it, Syntology harvested nothing from it; placed_by_identical_code counts samples whose link records no repository, placed here because identical code was harvested from this repository for another paper; ran counts executions on a synthesized input and instrument counts failures of Syntology's instrument, not of the code"},"code_links_note":null,"arxiv_metadata":null,"syntology_extracted_results":null,"record_sha256":"3ca45dbb166afd6a0b34075a048ac9d49582bd7351afb10a4e0880a21e510dc3","record_changed_at":"2026-09-28","record_changed_at_basis":"first_hashed"}