{"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/arxiv-2509-25050","title":"Advantage Weighted Matching: Aligning RL with Pretraining in Diffusion Models","arxiv_id":"2509.25050","date":"2025-09-29","proceeding":"ICML","authors":["Shuchen Xue","Chongjian Ge","Shilong Zhang","Yichen Li","Zhi-Ming Ma"],"abstract":"Reinforcement Learning (RL) has emerged as a central paradigm for advancing Large Language Models (LLMs), where both pre-training and RL post-training stages are grounded in the same log-likelihood formulation. In contrast, recent RL approaches for diffusion models, most notably Denoising Diffusion Policy Optimization (DDPO), optimize an objective different from the pretraining objectives--score/flow matching loss. In this work, we establish a novel theoretical analysis: DDPO is an implicit form of score/flow matching with noisy targets, which increases variance and slows convergence. Building on this analysis, we introduce Advantage Weighted Matching (AWM), a policy-gradient method for diffusion. It uses the score/flow-matching loss and reweights each sample by its advantage. In effect, AWM raises the influence of high-reward samples and suppresses low-reward ones while keeping the modeling objective identical to pretraining. This simple yet effective design yields substantial benefits: on the GenEval, OCR, and PickScore benchmarks, AWM delivers up to a $34\\times$ speedup over Flow-GRPO (which builds on DDPO), when applied to Stable Diffusion 3.5 Medium and FLUX, without compromising generation quality. Code is available at https://github.com/scxue/advantage_weighted_matching","url_abs":"https://arxiv.org/abs/2509.25050","url_pdf":"https://arxiv.org/pdf/2509.25050","source":{"archive":null,"snapshot":"2025-07-28","note":"not in the Papers with Code archive (frozen at the snapshot)","row_kind":"graph","title_abstract_authors_date":"arXiv metadata, CC0 1.0 (https://info.arxiv.org/help/license)"},"code_links":[],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2509.25050","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2509.25050"}},"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":[{"mentioned_in_github":null,"is_official":null,"provenance":"deterministic:regex_extraction","mentioned_in_paper":null,"url":"https://github.com/scxue/advantage_weighted_matching","reach":{"status":"ok","spdx":"Apache-2.0"}}],"summary":{"unverified":8},"by_repo_kind":{"found_in_text":{"samples":8,"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":"7558df3f898efd99","entry":"distribute_samples","repo":"scxue/advantage_weighted_matching","repo_kind":"found_in_text","path":"advantage_weighted_matching/dataset/merge_genevaltask.py","file_url":"https://github.com/scxue/advantage_weighted_matching/blob/HEAD/advantage_weighted_matching/dataset/merge_genevaltask.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"7558df3f898efd99"}},{"code_sha256_prefix":"aaf3538ea48a524d","entry":"extract_scores","repo":"scxue/advantage_weighted_matching","repo_kind":"found_in_text","path":"advantage_weighted_matching/flow_grpo/qwenvl.py","file_url":"https://github.com/scxue/advantage_weighted_matching/blob/HEAD/advantage_weighted_matching/flow_grpo/qwenvl.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"aaf3538ea48a524d"}},{"code_sha256_prefix":"b6e22297f43ef430","entry":"get_config","repo":"scxue/advantage_weighted_matching","repo_kind":"found_in_text","path":"advantage_weighted_matching/config/dgx_awm.py","file_url":"https://github.com/scxue/advantage_weighted_matching/blob/HEAD/advantage_weighted_matching/config/dgx_awm.py","link_basis":"plan_row","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"b6e22297f43ef430"}},{"code_sha256_prefix":"679c61d741aabb1d","entry":"pil_image_to_base64","repo":"scxue/advantage_weighted_matching","repo_kind":"found_in_text","path":"advantage_weighted_matching/flow_grpo/qwenvl.py","file_url":"https://github.com/scxue/advantage_weighted_matching/blob/HEAD/advantage_weighted_matching/flow_grpo/qwenvl.py","link_basis":"plan_row","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"679c61d741aabb1d"}},{"code_sha256_prefix":"2b06f4f15a325a0c","entry":"prompt_training_seed","repo":"scxue/advantage_weighted_matching","repo_kind":"found_in_text","path":"advantage_weighted_matching/flow_grpo/prompt_noise.py","file_url":"https://github.com/scxue/advantage_weighted_matching/blob/HEAD/advantage_weighted_matching/flow_grpo/prompt_noise.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"2b06f4f15a325a0c"}},{"code_sha256_prefix":"90a22cdf7d22703b","entry":"read_jsonl","repo":"scxue/advantage_weighted_matching","repo_kind":"found_in_text","path":"advantage_weighted_matching/dataset/merge_genevaltask.py","file_url":"https://github.com/scxue/advantage_weighted_matching/blob/HEAD/advantage_weighted_matching/dataset/merge_genevaltask.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"90a22cdf7d22703b"}},{"code_sha256_prefix":"eca464d584595994","entry":"sample_data","repo":"scxue/advantage_weighted_matching","repo_kind":"found_in_text","path":"advantage_weighted_matching/dataset/merge_genevaltask.py","file_url":"https://github.com/scxue/advantage_weighted_matching/blob/HEAD/advantage_weighted_matching/dataset/merge_genevaltask.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"eca464d584595994"}},{"code_sha256_prefix":"bfac543ac02f1dd0","entry":"shared_prompt_training_inputs","repo":"scxue/advantage_weighted_matching","repo_kind":"found_in_text","path":"advantage_weighted_matching/flow_grpo/prompt_noise.py","file_url":"https://github.com/scxue/advantage_weighted_matching/blob/HEAD/advantage_weighted_matching/flow_grpo/prompt_noise.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"bfac543ac02f1dd0"}}]},"arxiv_metadata":{"licence":"arXiv metadata, CC0 1.0 (https://info.arxiv.org/help/license)","fields":["title","abstract","authors","date"],"primary_category":"cs.LG","source":"arxiv_api"},"syntology_extracted_results":null}