{"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/aligning-few-step-diffusion-models-with-dense","title":"Aligning Few-Step Diffusion Models with Dense Reward Difference Learning","arxiv_id":"2411.11727","date":"2024-11-18","proceeding":null,"authors":["Ziyi Zhang","Li Shen","Sen Zhang","Deheng Ye","Yong Luo","Miaojing Shi","Bo Du","DaCheng Tao"],"abstract":"Aligning diffusion models with downstream objectives is essential for their practical applications. However, standard alignment methods often struggle with step generalization when directly applied to few-step diffusion models, leading to inconsistent performance across different denoising step scenarios. To address this, we introduce Stepwise Diffusion Policy Optimization (SDPO), a novel alignment method tailored for few-step diffusion models. Unlike prior approaches that rely on a single sparse reward from only the final step of each denoising trajectory for trajectory-level optimization, SDPO incorporates dense reward feedback at every intermediate step. By learning the differences in dense rewards between paired samples, SDPO facilitates stepwise optimization of few-step diffusion models, ensuring consistent alignment across all denoising steps. To promote stable and efficient training, SDPO introduces an online reinforcement learning framework featuring several novel strategies designed to effectively exploit the stepwise granularity of dense rewards. Experimental results demonstrate that SDPO consistently outperforms prior methods in reward-based alignment across diverse step configurations, underscoring its robust step generalization capabilities. Code is avaliable at https://github.com/ZiyiZhang27/sdpo.","url_abs":"https://arxiv.org/abs/2411.11727v1","url_pdf":"https://arxiv.org/pdf/2411.11727v1.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":"aligning-few-step-diffusion-models-with-dense","repo_url":"https://github.com/ziyizhang27/sdpo","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"denoising","task_name":"Denoising"}],"methods":[{"method_slug":"diffusion","method_name":"Diffusion"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2411.11727","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2411.11727"}},"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/ziyizhang27/sdpo","reach":null}],"summary":{"unverified":4},"by_repo_kind":{"official":{"samples":4,"ran":0,"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":0,"samples":[{"code_sha256_prefix":"b47084c4522e837b","entry":"ImageReward","repo":"ziyizhang27/sdpo","repo_kind":"official","path":"sdpo_pytorch/rewards.py","file_url":"https://github.com/ziyizhang27/sdpo/blob/HEAD/sdpo_pytorch/rewards.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":"b47084c4522e837b"}},{"code_sha256_prefix":"d0dd9ffc825abaee","entry":"aesthetic_score","repo":"ziyizhang27/sdpo","repo_kind":"official","path":"sdpo_pytorch/rewards.py","file_url":"https://github.com/ziyizhang27/sdpo/blob/HEAD/sdpo_pytorch/rewards.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":"d0dd9ffc825abaee"}},{"code_sha256_prefix":"b6e22297f43ef430","entry":"get_config","repo":"ziyizhang27/sdpo","repo_kind":"official","path":"config/config_sdpo.py","file_url":"https://github.com/ziyizhang27/sdpo/blob/HEAD/config/config_sdpo.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":"b6e22297f43ef430"}},{"code_sha256_prefix":"e1a404ba4ab6e57a","entry":"hpsv2","repo":"ziyizhang27/sdpo","repo_kind":"official","path":"sdpo_pytorch/rewards.py","file_url":"https://github.com/ziyizhang27/sdpo/blob/HEAD/sdpo_pytorch/rewards.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":"e1a404ba4ab6e57a"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}