{"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/personalized-soups-personalized-large","title":"Personalized Soups: Personalized Large Language Model Alignment via Post-hoc Parameter Merging","arxiv_id":"2310.11564","date":"2023-10-17","proceeding":null,"authors":["Joel Jang","Seungone Kim","Bill Yuchen Lin","Yizhong Wang","Jack Hessel","Luke Zettlemoyer","Hannaneh Hajishirzi","Yejin Choi","Prithviraj Ammanabrolu"],"abstract":"While Reinforcement Learning from Human Feedback (RLHF) aligns Large Language Models (LLMs) with general, aggregate human preferences, it is suboptimal for learning diverse, individual perspectives. In this work, we study Reinforcement Learning from Personalized Human Feedback (RLPHF) problem, wherein LLMs are aligned to multiple (sometimes conflicting) preferences by modeling alignment as a Multi-Objective Reinforcement Learning (MORL) problem. Compared to strong single-objective baselines, we show that we can achieve personalized alignment by decomposing preferences into multiple dimensions. These dimensions are defined based on personalizations that are declared as desirable by the user. In this work, we show that they can be efficiently trained independently in a distributed manner and combined effectively post-hoc through parameter merging. The code is available at https://github.com/joeljang/RLPHF.","url_abs":"https://arxiv.org/abs/2310.11564v1","url_pdf":"https://arxiv.org/pdf/2310.11564v1.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":"personalized-soups-personalized-large","repo_url":"https://github.com/joeljang/rlphf","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}}],"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"},{"task_slug":"multi-objective-reinforcement-learning","task_name":"Multi-Objective Reinforcement Learning"},{"task_slug":"reinforcement-learning","task_name":"Reinforcement Learning"},{"task_slug":"reinforcement-learning-2","task_name":"reinforcement-learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2310.11564","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2310.11564"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+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/joeljang/rlphf","reach":{"status":"ok"}}],"summary":{"ran":12,"ran_draft_wrong":1,"unverified":2},"by_repo_kind":{"official":{"samples":15,"ran":13,"repositories":1}},"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":15,"samples":[{"code_sha256_prefix":"ec7ad96cced15dae","entry":"chars_token_ratio","repo":"joeljang/rlphf","repo_kind":"official","path":"training/multitask_training.py","file_url":"https://github.com/joeljang/rlphf/blob/HEAD/training/multitask_training.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"ec7ad96cced15dae"}},{"code_sha256_prefix":"bda36ad0c44dbd74","entry":"collator","repo":"joeljang/rlphf","repo_kind":"official","path":"training/pmorl.py","file_url":"https://github.com/joeljang/rlphf/blob/HEAD/training/pmorl.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"bda36ad0c44dbd74"}},{"code_sha256_prefix":"cca45bb9a111ab77","entry":"compute_reward_modeling_metrics","repo":"joeljang/rlphf","repo_kind":"official","path":"gpt4_evaluate/alpaca_farm/reward_modeling_trainer.py","file_url":"https://github.com/joeljang/rlphf/blob/HEAD/gpt4_evaluate/alpaca_farm/reward_modeling_trainer.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"cca45bb9a111ab77"}},{"code_sha256_prefix":"f5ff77db7757bc92","entry":"flatten_dict","repo":"joeljang/rlphf","repo_kind":"official","path":"gpt4_evaluate/alpaca_farm/common.py","file_url":"https://github.com/joeljang/rlphf/blob/HEAD/gpt4_evaluate/alpaca_farm/common.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"f5ff77db7757bc92"}},{"code_sha256_prefix":"46af3eb303562401","entry":"format_output","repo":"joeljang/rlphf","repo_kind":"official","path":"gpt4_evaluate/alpaca_farm/data_preprocessor.py","file_url":"https://github.com/joeljang/rlphf/blob/HEAD/gpt4_evaluate/alpaca_farm/data_preprocessor.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"46af3eb303562401"}},{"code_sha256_prefix":"a71a4bdc37586fe2","entry":"format_prompt","repo":"joeljang/rlphf","repo_kind":"official","path":"gpt4_evaluate/alpaca_farm/data_preprocessor.py","file_url":"https://github.com/joeljang/rlphf/blob/HEAD/gpt4_evaluate/alpaca_farm/data_preprocessor.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"a71a4bdc37586fe2"}},{"code_sha256_prefix":"8b1ea4142642d361","entry":"format_prompt_with_data_frame","repo":"joeljang/rlphf","repo_kind":"official","path":"gpt4_evaluate/alpaca_farm/data_preprocessor.py","file_url":"https://github.com/joeljang/rlphf/blob/HEAD/gpt4_evaluate/alpaca_farm/data_preprocessor.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"8b1ea4142642d361"}},{"code_sha256_prefix":"dad9dcf30ea1d344","entry":"generate_prompt","repo":"joeljang/rlphf","repo_kind":"official","path":"generate_rollouts.py","file_url":"https://github.com/joeljang/rlphf/blob/HEAD/generate_rollouts.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"dad9dcf30ea1d344"}},{"code_sha256_prefix":"6d93a251945458d6","entry":"generate_prompt_tulu","repo":"joeljang/rlphf","repo_kind":"official","path":"generate_rollouts.py","file_url":"https://github.com/joeljang/rlphf/blob/HEAD/generate_rollouts.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"6d93a251945458d6"}},{"code_sha256_prefix":"ed46234879c19e1b","entry":"get_reward_stats","repo":"joeljang/rlphf","repo_kind":"official","path":"training/pmorl.py","file_url":"https://github.com/joeljang/rlphf/blob/HEAD/training/pmorl.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"ed46234879c19e1b"}},{"code_sha256_prefix":"9a1871dd5c967ac1","entry":"let_model_save_mem_when_zero_grad","repo":"joeljang/rlphf","repo_kind":"official","path":"gpt4_evaluate/alpaca_farm/common.py","file_url":"https://github.com/joeljang/rlphf/blob/HEAD/gpt4_evaluate/alpaca_farm/common.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"9a1871dd5c967ac1"}},{"code_sha256_prefix":"c91f5610ae8e27c8","entry":"prepare_sample_text","repo":"joeljang/rlphf","repo_kind":"official","path":"training/multitask_training.py","file_url":"https://github.com/joeljang/rlphf/blob/HEAD/training/multitask_training.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"c91f5610ae8e27c8"}},{"code_sha256_prefix":"5240cc5f3702d81a","entry":"preprocess_function","repo":"joeljang/rlphf","repo_kind":"official","path":"training_reward_model.py","file_url":"https://github.com/joeljang/rlphf/blob/HEAD/training_reward_model.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"5240cc5f3702d81a"}},{"code_sha256_prefix":"fb96ca75dd81e5fc","entry":"add_preference_prompt_during_val","repo":"joeljang/rlphf","repo_kind":"official","path":"training/psoups.py","file_url":"https://github.com/joeljang/rlphf/blob/HEAD/training/psoups.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"fb96ca75dd81e5fc"}},{"code_sha256_prefix":"fc2e047ee0f8e902","entry":"build_dataset","repo":"joeljang/rlphf","repo_kind":"official","path":"training/pmorl.py","file_url":"https://github.com/joeljang/rlphf/blob/HEAD/training/pmorl.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"fc2e047ee0f8e902"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}