{"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/plan-for-speed-dilated-scheduling-for-masked","title":"Plan for Speed -- Dilated Scheduling for Masked Diffusion Language Models","arxiv_id":"2506.19037","date":"2025-06-23","proceeding":null,"authors":["Omer Luxembourg","Haim Permuter","Eliya Nachmani"],"abstract":"Masked diffusion language models (MDLM) have shown strong promise for non-autoregressive text generation, yet existing samplers act as implicit planners, selecting tokens to unmask via denoiser confidence or entropy scores. Such heuristics falter under parallel unmasking - they ignore pairwise interactions between tokens and cannot account for dependencies when unmasking multiple positions at once, limiting their inference time to traditional auto-regressive (AR) models. We introduce the Dilated-scheduled Unmasking Strategy (DUS), an inference-only, planner-model-free method that requires no additional training. DUS leverages a first-order Markov assumption to partition sequence positions into dilation-based groups of non-adjacent tokens, enabling independent, parallel unmasking steps that respect local context that minimizes the joint entropy of each iteration step. Unlike semi-AR block approaches (e.g., LLADA and Dream) that still invoke the denoiser per block, DUS reduces the number of denoiser calls to O(log B) per generation block - yielding substantial speedup over the O(B) run time of state-of-the-art diffusion models, where B is the block size in the semi-AR inference process. In experiments on math (GSM8K) and code completion (Humaneval, MBPP) benchmarks - domains suited to non-ordinal generation - DUS improves scores over parallel confidence-based planner, without modifying the underlying denoiser. DUS offers a lightweight, budget-aware approach to efficient, high-quality text generation, paving the way to unlock the true capabilities of MDLMs.","url_abs":"https://arxiv.org/abs/2506.19037v1","url_pdf":"https://arxiv.org/pdf/2506.19037v1.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":[],"tasks":[{"task_slug":"code-completion","task_name":"Code Completion"},{"task_slug":"gsm8k","task_name":"GSM8K"},{"task_slug":"humaneval","task_name":"HumanEval"},{"task_slug":"math","task_name":"Math"},{"task_slug":"scheduling","task_name":"Scheduling"},{"task_slug":"text-generation","task_name":"Text Generation"},{"task_slug":null,"task_name":"mbpp"}],"methods":[{"method_slug":"diffusion","method_name":"Diffusion"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2506.19037","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2506.19037"}},"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":"deterministic:regex_extraction","url":"https://github.com/omerlux/DUS","reach":null},{"provenance":"deterministic:regex_extraction","url":"https://github.com/EleutherAI/lm-eval-harness","reach":{"status":"gone","observed_at":"2026-09-16","how":"tree_404+repo_404"}},{"provenance":"deterministic:regex_extraction","url":"https://github.com/ML-GSAI/LLaDA","reach":null}],"summary":{"ran_fixture":2,"ran_violates":1,"ran_draft_wrong":1,"unverified":1},"by_repo_kind":{"found_in_text":{"samples":4,"ran":3,"repositories":2}},"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":4,"samples":[{"code_sha256_prefix":"b8d1edd147098bad","entry":"apply_end_think_logit_boost","repo":"ML-GSAI/LLaDA","repo_kind":"found_in_text","path":"generate.py","file_url":"https://github.com/ML-GSAI/LLaDA/blob/HEAD/generate.py","link_basis":"first_harvest_node","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"b8d1edd147098bad"}},{"code_sha256_prefix":"f3fec99b5aa1aa8e","entry":"contains_token_sequence","repo":null,"repo_kind":null,"path":null,"file_url":null,"link_basis":"identical_code_first_harvested_elsewhere","language":"python","status":"ran_violates","verification_level":1,"contract_check":"VIOLATES","metamorphic_tier":"deterministic","behaviour_fingerprint":true,"licence":null,"inline_ok":false,"mcp_get_code":{"code_sha256":"f3fec99b5aa1aa8e"}},{"code_sha256_prefix":"9d2036ccca0be595","entry":"dilated_unmask_levels","repo":"omerlux/DUS","repo_kind":"found_in_text","path":"generate.py","file_url":"https://github.com/omerlux/DUS/blob/HEAD/generate.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"9d2036ccca0be595"}},{"code_sha256_prefix":"94947c39a3c54f82","entry":"get_next_sequence_token_id","repo":"ML-GSAI/LLaDA","repo_kind":"found_in_text","path":"generate.py","file_url":"https://github.com/ML-GSAI/LLaDA/blob/HEAD/generate.py","link_basis":"first_harvest_node","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"94947c39a3c54f82"}},{"code_sha256_prefix":"6f9738ad733e164b","entry":"generate","repo":"ML-GSAI/LLaDA","repo_kind":"found_in_text","path":"generate.py","file_url":"https://github.com/ML-GSAI/LLaDA/blob/HEAD/generate.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":"6f9738ad733e164b"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}