{"about":{"site":"https://codewithpapers.app","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.","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"},"url":"/paper/discovering-preference-optimization","title":"Discovering Preference Optimization Algorithms with and for Large Language Models","arxiv_id":"2406.08414","date":"2024-06-12","proceeding":null,"authors":["Chris Lu","Samuel Holt","Claudio Fanconi","Alex J. Chan","Jakob Foerster","Mihaela van der Schaar","Robert Tjarko Lange"],"abstract":"Offline preference optimization is a key method for enhancing and controlling the quality of Large Language Model (LLM) outputs. Typically, preference optimization is approached as an offline supervised learning task using manually-crafted convex loss functions. While these methods are based on theoretical insights, they are inherently constrained by human creativity, so the large search space of possible loss functions remains under explored. We address this by performing LLM-driven objective discovery to automatically discover new state-of-the-art preference optimization algorithms without (expert) human intervention. Specifically, we iteratively prompt an LLM to propose and implement new preference optimization loss functions based on previously-evaluated performance metrics. This process leads to the discovery of previously-unknown and performant preference optimization algorithms. The best performing of these we call Discovered Preference Optimization (DiscoPOP), a novel algorithm that adaptively blends logistic and exponential losses. Experiments demonstrate the state-of-the-art performance of DiscoPOP and its successful transfer to held-out tasks.","url_abs":"https://arxiv.org/abs/2406.08414v3","url_pdf":"https://arxiv.org/pdf/2406.08414v3.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":"discovering-preference-optimization","repo_url":"https://github.com/luchris429/DiscoPOP","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"discovering-preference-optimization","repo_url":"https://github.com/samholt/DiscoPOP","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"discovering-preference-optimization","repo_url":"https://github.com/vanderschaarlab/discopop","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"language-modeling","task_name":"Language Modeling"},{"task_slug":"language-modelling","task_name":"Language Modelling"},{"task_slug":"large-language-model","task_name":"Large Language Model"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/2406.08414","atlas_url":"https://app.syntology.ai/?focus=2406.08414","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2406.08414"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-25T09:33:49+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/vanderschaarlab/discopop","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/luchris429/DiscoPOP","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/samholt/DiscoPOP","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"ran":4,"ran_draft_wrong":2,"ran_violates":1,"unverified":1},"by_repo_kind":{"official":{"samples":8,"ran":7,"repositories":2}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"9be97bcdb51fd634","entry":"apply_chat_template","repo":"luchris429/DiscoPOP","repo_kind":"official","path":"src/alignment/data.py","file_url":"https://github.com/luchris429/DiscoPOP/blob/HEAD/src/alignment/data.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"9be97bcdb51fd634"}},{"code_sha256_prefix":"0edeacd7a9034f53","entry":"extract_docstring","repo":"luchris429/DiscoPOP","repo_kind":"official","path":"src/alignment/decontaminate.py","file_url":"https://github.com/luchris429/DiscoPOP/blob/HEAD/src/alignment/decontaminate.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"0edeacd7a9034f53"}},{"code_sha256_prefix":"96f002b17b0af96e","entry":"init_archive","repo":"vanderschaarlab/discopop","repo_kind":"official","path":"scripts/launch_evo.py","file_url":"https://github.com/vanderschaarlab/discopop/blob/HEAD/scripts/launch_evo.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"96f002b17b0af96e"}},{"code_sha256_prefix":"4253aac010c4504f","entry":"is_openai_format","repo":"luchris429/DiscoPOP","repo_kind":"official","path":"src/alignment/data.py","file_url":"https://github.com/luchris429/DiscoPOP/blob/HEAD/src/alignment/data.py","link_basis":"harvester_set","language":"python","status":"ran_violates","verification_level":1,"contract_check":"VIOLATES","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"4253aac010c4504f"}},{"code_sha256_prefix":"a3791c5c2185e013","entry":"load_dataset_column","repo":"luchris429/DiscoPOP","repo_kind":"official","path":"src/alignment/decontaminate.py","file_url":"https://github.com/luchris429/DiscoPOP/blob/HEAD/src/alignment/decontaminate.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"a3791c5c2185e013"}},{"code_sha256_prefix":"f43e59d239019e86","entry":"normalize_whitespace","repo":"luchris429/DiscoPOP","repo_kind":"official","path":"src/alignment/decontaminate.py","file_url":"https://github.com/luchris429/DiscoPOP/blob/HEAD/src/alignment/decontaminate.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"f43e59d239019e86"}},{"code_sha256_prefix":"6c1e6f76c77c843d","entry":"validate_code","repo":"vanderschaarlab/discopop","repo_kind":"official","path":"scripts/launch_evo.py","file_url":"https://github.com/vanderschaarlab/discopop/blob/HEAD/scripts/launch_evo.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"6c1e6f76c77c843d"}},{"code_sha256_prefix":"d303691c35493431","entry":"train_gpo","repo":"vanderschaarlab/discopop","repo_kind":"official","path":"scripts/launch_evo.py","file_url":"https://github.com/vanderschaarlab/discopop/blob/HEAD/scripts/launch_evo.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"d303691c35493431"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}