.</p>\n","updatedAt":"2026-08-13T06:54:35.557Z","author":{"_id":"652066649004117947e46ed6","avatarUrl":"/avatars/972c97df6f26d2c3d6ce71ec579984bb.svg","fullname":"Jaehong Yoon","name":"jaehong31","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":5,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"fr","probability":0.32275810837745667},"editors":["jaehong31"],"editorAvatarUrls":["/avatars/972c97df6f26d2c3d6ce71ec579984bb.svg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2608.10615","authors":[{"_id":"6a7d6a0c0ac8bee77474ef90","name":"Jinya Sakurai","hidden":false},{"_id":"6a7d6a0c0ac8bee77474ef91","name":"Patrick Pynadath","hidden":false},{"_id":"6a7d6a0c0ac8bee77474ef92","name":"Satoshi Hayakawa","hidden":false},{"_id":"6a7d6a0c0ac8bee77474ef93","name":"Jaehong Yoon","hidden":false},{"_id":"6a7d6a0c0ac8bee77474ef94","name":"Xulei Yang","hidden":false},{"_id":"6a7d6a0c0ac8bee77474ef95","name":"Nancy F. Chen","hidden":false},{"_id":"6a7d6a0c0ac8bee77474ef96","name":"Xun Xu","hidden":false}],"publishedAt":"2026-08-11T00:00:00.000Z","submittedOnDailyAt":"2026-08-13T00:00:00.000Z","title":"Simplex Relaxation for Discrete Diffusion","submittedOnDailyBy":{"_id":"652066649004117947e46ed6","avatarUrl":"/avatars/972c97df6f26d2c3d6ce71ec579984bb.svg","isPro":false,"fullname":"Jaehong Yoon","user":"jaehong31","type":"user","name":"jaehong31"},"summary":"Discrete diffusion models for categorical generation are defined by a corruption kernel, which determines the intermediate state space and the associated reverse prediction problem. We study uniform discrete diffusion and ask whether its training objective and reverse transitions can be enriched without changing the underlying categorical corruption process. We introduce Simplax, an exact Dirichlet--categorical augmentation that couples each corrupted categorical state with an auxiliary simplex-valued variable while preserving the original uniform diffusion process as its categorical marginal. This augmentation yields a tractable Rao--Blackwellized reverse-bridge objective and a corresponding stochastic reverse sampler, while retaining the corrupted categorical state as the denoiser input. Empirically, Simplax improves the generative perplexity--entropy tradeoff on unconditional OpenWebText generation. On Sudoku, a model trained exclusively on 30-clue puzzles achieves the highest accuracy among the compared methods across all evaluated clue densities, including the minimum uniquely solvable 17-clue regime, and also achieves the highest validity in unconditional generation.","upvotes":2,"discussionId":"6a7d6a0c0ac8bee77474ef97","ai_summary":"Simplax enriches uniform discrete diffusion via Dirichlet-categorical augmentation to improve reverse sampling and generative quality on text and Sudoku tasks.","ai_keywords":["discrete diffusion models","corruption kernel","uniform discrete diffusion","Dirichlet-categorical augmentation","Rao-Blackwellized reverse-bridge objective","stochastic reverse sampler","generative perplexity-entropy tradeoff"],"ai_summary_model":"thinkingmachines/Inkling-Small"},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"652066649004117947e46ed6","avatarUrl":"/avatars/972c97df6f26d2c3d6ce71ec579984bb.svg","isPro":false,"fullname":"Jaehong Yoon","user":"jaehong31","type":"user"},{"_id":"678604dea34abb89c6920834","avatarUrl":"/avatars/45f2c4a718c49cfcfa944ef78205674f.svg","isPro":false,"fullname":"SeungBum Ha","user":"SeungB","type":"user"}],"acceptLanguages":["en"],"dailyPaperRank":0,"markdownContentUrl":"https://huggingface.co/buckets/huggingchat/papers-content/resolve/2608/2608.10615.md","query":{}}">
Simplex Relaxation for Discrete Diffusion
Abstract
Simplax enriches uniform discrete diffusion via Dirichlet-categorical augmentation to improve reverse sampling and generative quality on text and Sudoku tasks.
Discrete diffusion models for categorical generation are defined by a corruption kernel, which determines the intermediate state space and the associated reverse prediction problem. We study uniform discrete diffusion and ask whether its training objective and reverse transitions can be enriched without changing the underlying categorical corruption process. We introduce Simplax, an exact Dirichlet--categorical augmentation that couples each corrupted categorical state with an auxiliary simplex-valued variable while preserving the original uniform diffusion process as its categorical marginal. This augmentation yields a tractable Rao--Blackwellized reverse-bridge objective and a corresponding stochastic reverse sampler, while retaining the corrupted categorical state as the denoiser input. Empirically, Simplax improves the generative perplexity--entropy tradeoff on unconditional OpenWebText generation. On Sudoku, a model trained exclusively on 30-clue puzzles achieves the highest accuracy among the compared methods across all evaluated clue densities, including the minimum uniquely solvable 17-clue regime, and also achieves the highest validity in unconditional generation.
Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.
Tap or paste here to upload images
Discussion (0)
Sign in to join the discussion. Free account, 30 seconds — email code or GitHub.
Sign in →No comments yet. Sign in and be the first to say something.