Z-Image-Pixel &amp; Empirical Insight of Training Pixel-Space Diffusion Models</p>\n","updatedAt":"2026-08-18T03:30:29.862Z","author":{"_id":"662a0f2d4bab737c1a279843","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/662a0f2d4bab737c1a279843/fC2p3mjMHkVpDQdEqkuR4.png","fullname":"Dengyang Jiang","name":"DyJiang","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":16,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.3316993713378906},"editors":["DyJiang"],"editorAvatarUrls":["https://cdn-avatars.huggingface.co/v1/production/uploads/662a0f2d4bab737c1a279843/fC2p3mjMHkVpDQdEqkuR4.png"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2608.16887","authors":[{"_id":"6a83d19d675db694db8cd573","user":{"_id":"662a0f2d4bab737c1a279843","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/662a0f2d4bab737c1a279843/fC2p3mjMHkVpDQdEqkuR4.png","isPro":false,"fullname":"Dengyang Jiang","user":"DyJiang","type":"user","name":"DyJiang"},"name":"Dengyang Jiang","status":"claimed_verified","statusLastChangedAt":"2026-08-18T08:45:05.191Z","hidden":false},{"_id":"6a83d19d675db694db8cd574","name":"Ruoyi Du","hidden":false},{"_id":"6a83d19d675db694db8cd575","name":"Zhennan Chen","hidden":false},{"_id":"6a83d19d675db694db8cd576","name":"Dongyang Liu","hidden":false},{"_id":"6a83d19d675db694db8cd577","name":"Zanyi Wang","hidden":false},{"_id":"6a83d19d675db694db8cd578","name":"Mingzhe Zheng","hidden":false},{"_id":"6a83d19d675db694db8cd579","name":"Xiangpeng Yang","hidden":false},{"_id":"6a83d19d675db694db8cd57a","name":"Huanqia Cai","hidden":false},{"_id":"6a83d19d675db694db8cd57b","name":"Aiming Hao","hidden":false},{"_id":"6a83d19d675db694db8cd57c","name":"Yuming Jiang","hidden":false},{"_id":"6a83d19d675db694db8cd57d","name":"Peng Gao","hidden":false},{"_id":"6a83d19d675db694db8cd57e","name":"Harry Yang","hidden":false},{"_id":"6a83d19d675db694db8cd57f","name":"Steven Hoi","hidden":false}],"publishedAt":"2026-08-17T00:00:00.000Z","submittedOnDailyAt":"2026-08-18T00:00:00.000Z","title":"An Empirical Study of Training Pixel-Space Text-to-Image Diffusion Models","submittedOnDailyBy":{"_id":"662a0f2d4bab737c1a279843","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/662a0f2d4bab737c1a279843/fC2p3mjMHkVpDQdEqkuR4.png","isPro":false,"fullname":"Dengyang Jiang","user":"DyJiang","type":"user","name":"DyJiang"},"summary":"This paper investigates an increasingly important topic in generative modeling: pixel-space diffusion models. Although numerous studies have explored this topic, most focus on small-scale or class-conditional settings. Consequently, a practical recipe for training pixel-space models that rival or exceed well-established latent-space counterparts remains elusive. Through a comprehensive empirical study, we first observe that direct large-scale pre-training in pixel space converges substantially more slowly than in latent space. This observation motivates a latent-to-pixel strategy that acquires generative priors efficiently in latent space and transitions to pixel space during post-training. We then systematically investigate the key design choices governing this transition, including weight initialization, data composition, prediction target, decoder architecture, and noise schedule, and identify a practical recipe that makes the resulting pixel-space models match or outperform their latent-space counterparts while delivering 3.18 to 4.75 times end-to-end inference speedups. We hope that our findings provide useful empirical insights and practical guidelines for future research on pixel-space generation.","upvotes":19,"discussionId":"6a83d19e675db694db8cd580","ai_summary":"Researchers propose a latent-to-pixel training strategy that accelerates convergence and improves inference speed for large-scale pixel-space diffusion models.","ai_keywords":["pixel-space diffusion models","latent-to-pixel strategy","generative priors","weight initialization","prediction target","decoder architecture","noise schedule"],"ai_summary_model":"thinkingmachines/Inkling-Small","organization":{"_id":"6925b20fed452d1567c012d3","name":"Tongyi-MAI","fullname":"Tongyi-MAI","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/64379d79fac5ea753f1c10f3/fxHO6QoYjdv9_LTyiUD3g.jpeg"}},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"662a0f2d4bab737c1a279843","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/662a0f2d4bab737c1a279843/fC2p3mjMHkVpDQdEqkuR4.png","isPro":false,"fullname":"Dengyang Jiang","user":"DyJiang","type":"user"},{"_id":"65f2e3d1cec22d29ce41ef94","avatarUrl":"/avatars/a724073aa1ae5d72847955d2f39773fa.svg","isPro":true,"fullname":"Zanyi Wang","user":"xmz111","type":"user"},{"_id":"6570450a78d7aca0c361a177","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6570450a78d7aca0c361a177/MX7jHhTQwLs-BvYIu5rqb.jpeg","isPro":false,"fullname":"Harold Chen","user":"Harold328","type":"user"},{"_id":"66449e619ff401732687f013","avatarUrl":"/avatars/251897d1324a70a9bf761513871c5841.svg","isPro":false,"fullname":"chen","user":"zhen-nan","type":"user"},{"_id":"6486df66373f79a52913e017","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6486df66373f79a52913e017/vUncXohxJN4ixXR6QUMxh.jpeg","isPro":true,"fullname":"Xiangpeng Yang","user":"XiangpengYang","type":"user"},{"_id":"62f0c4abe2999b231e5a893c","avatarUrl":"/avatars/90da268b877a7ffe6665075c84018a83.svg","isPro":false,"fullname":"Mingzhe Zheng","user":"Dunge0nMaster","type":"user"},{"_id":"620783f24e28382272337ba4","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/620783f24e28382272337ba4/zkUveQPNiDfYjgGhuFErj.jpeg","isPro":false,"fullname":"GuoLiangTang","user":"Tommy930","type":"user"},{"_id":"637c22183d8e2e9c40c09fcf","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/1669079538761-noauth.jpeg","isPro":false,"fullname":"Zhizhou Chen","user":"Chenzzzzzz","type":"user"},{"_id":"646f1bef075e11ca78da3bb7","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/646f1bef075e11ca78da3bb7/gNS-ikyZXYeMrf4a7HTQE.jpeg","isPro":false,"fullname":"Dongyang Liu (Chris Liu)","user":"Cxxs","type":"user"},{"_id":"651f8133dbf879b8c58f5136","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/651f8133dbf879b8c58f5136/0L8Ecgi5Ietkm_DchJwE-.png","isPro":false,"fullname":"Zikai Zhou","user":"Klayand","type":"user"},{"_id":"629c95b7a5d6f5fe10e6ed45","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/629c95b7a5d6f5fe10e6ed45/Sy0Ype5snsRookID-gsSm.jpeg","isPro":false,"fullname":"Yuming Jiang","user":"yumingj","type":"user"},{"_id":"6589b61dbfdd9f4410af9b7d","avatarUrl":"/avatars/7d644e750f6084d4f24e332135cc5be8.svg","isPro":false,"fullname":"Hu","user":"Irving1","type":"user"}],"acceptLanguages":["en"],"dailyPaperRank":0,"organization":{"_id":"6925b20fed452d1567c012d3","name":"Tongyi-MAI","fullname":"Tongyi-MAI","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/64379d79fac5ea753f1c10f3/fxHO6QoYjdv9_LTyiUD3g.jpeg"},"markdownContentUrl":"https://huggingface.co/buckets/huggingchat/papers-content/resolve/2608/2608.16887.md","query":{}}">

Get this paper in your agent:

Don't have the latest CLI?

Cite arxiv.org/abs/2608.16887 in a model README.md to link it from this page.

Cite arxiv.org/abs/2608.16887 in a dataset README.md to link it from this page.

Cite arxiv.org/abs/2608.16887 in a Space README.md to link it from this page.

Add this paper to a collection to link it from this page.