{"paper":{"title":"Free Draft-and-Verification: Toward Lossless Parallel Decoding for Diffusion Large Language Models","arxiv_id":"2510.00294","primary_cat":"cs.LG","submitted_at":"2025-09-30T21:28:04+00:00","version_compiled":"v4"},"notice":"Every quoted sentence is the authors' text from the arXiv listing abstract. Labels and probabilities are the outputs of classifier-1.13.0 on the question bank abstract-p1-20260923, exactly as returned. No sentence in this record was written by a language model.","schema":"pith.compiled-record/v0","status":"compiled","abstract":{"source":"arXiv listing abstract v4","sentences":[{"i":0,"p":1.0,"id":"s0","own":false,"role":"background","text":"Diffusion Large Language Models (dLLMs) have emerged as a compelling paradigm for language modeling, offering the unique potential for highly efficient inference via multi-token parallel decoding.","hedge":{"p":0.65,"label":"asserted"},"own_p":0.06,"main_claim_p":0.03,"quantitative":false,"quantitative_p":0.06,"role_probabilities":{"gap":0.0,"method":0.0,"result":0.0,"background":1.0,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.0}},{"i":1,"p":0.58,"id":"s1","own":false,"role":"gap","text":"However, unlocking this potential remains challenging.","hedge":{"p":0.98,"label":"asserted"},"own_p":0.05,"main_claim_p":0.02,"quantitative":false,"quantitative_p":0.02,"role_probabilities":{"gap":0.58,"method":0.0,"result":0.0,"background":0.41,"limitation":0.01,"future_work":0.0,"contribution":0.0,"interpretation":0.0}},{"i":2,"p":0.57,"id":"s2","own":false,"role":"background","text":"High-quality generation typically defaults to one-token-per-step static decoding, while existing parallel algorithms rely on heuristic confidence thresholds, introducing a fragile trade-off between decoding efficiency and generation quality.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.06,"main_claim_p":0.03,"quantitative":true,"quantitative_p":0.53,"role_probabilities":{"gap":0.43,"method":0.0,"result":0.0,"background":0.57,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.0}},{"i":3,"p":0.98,"id":"s3","own":true,"role":"contribution","text":"To overcome this, we introduce Free Draft-and-Verification (FreeDave), a novel training-free and model-free fast decoding algorithm.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.92,"main_claim_p":0.1,"quantitative":false,"quantitative_p":0.06,"role_probabilities":{"gap":0.0,"method":0.02,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.98,"interpretation":0.0}},{"i":4,"p":0.96,"id":"s4","own":true,"role":"method","text":"FreeDave capitalizes on the inherent properties of dLLMs by treating future token predictions as ``free drafts'' and rigorously self-verifying them in subsequent forward passes.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.8,"main_claim_p":0.07,"quantitative":false,"quantitative_p":0.04,"role_probabilities":{"gap":0.0,"method":0.96,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.04,"interpretation":0.0}},{"i":5,"p":0.51,"id":"s5","own":true,"role":"method","text":"Theoretically, FreeDave searches for a shorter decoding path that is guaranteed to reproduce the exact sequence generated by static decoding, achieving algorithmically lossless acceleration.","hedge":{"p":0.98,"label":"asserted"},"own_p":0.81,"main_claim_p":0.35,"quantitative":false,"quantitative_p":0.24,"role_probabilities":{"gap":0.0,"method":0.51,"result":0.02,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.36,"interpretation":0.11}},{"i":6,"p":0.64,"id":"s6","own":true,"role":"method","text":"With a sufficiently large draft budget and a verification frontier that does not retreat, it can use the fewest possible model forward passes among feasible paths.","hedge":{"p":0.6,"label":"asserted"},"own_p":0.65,"main_claim_p":0.07,"quantitative":false,"quantitative_p":0.43,"role_probabilities":{"gap":0.0,"method":0.64,"result":0.01,"background":0.0,"limitation":0.03,"future_work":0.0,"contribution":0.15,"interpretation":0.17}},{"i":7,"p":1.0,"id":"s7","own":true,"role":"result","text":"Extensive evaluations on math reasoning and code generation benchmarks demonstrate that FreeDave can boost the number of generated tokens per model forward (TPF) by up to 5.48$\\times$ and tokens per second (TPS) by up to 3.94$\\times$, while robustly maintaining generation quality and avoiding the task-specific performance degradation faced by heuristic methods.","hedge":{"p":0.99,"label":"asserted"},"own_p":0.95,"main_claim_p":0.77,"quantitative":true,"quantitative_p":0.99,"role_probabilities":{"gap":0.0,"method":0.0,"result":1.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.0}},{"i":8,"p":0.75,"id":"s8","own":false,"role":"contribution","text":"Code is available at https://github.com/cychomatica/FreeDave.","hedge":{"p":0.91,"label":"asserted"},"own_p":0.41,"main_claim_p":0.02,"quantitative":false,"quantitative_p":0.03,"role_probabilities":{"gap":0.0,"method":0.11,"result":0.02,"background":0.02,"limitation":0.0,"future_work":0.1,"contribution":0.75,"interpretation":0.0}}]},"compiled":{"main_claim":{"p":0.96,"text":"Extensive evaluations on math reasoning and code generation benchmarks demonstrate that FreeDave can boost the number of generated tokens per model forward (TPF) by up to 5.48$\\times$ and tokens per second (TPS) by up to 3.94$\\times$, while robustly maintaining generation quality and avoiding the task-specific performance degradation faced by heuristic methods.","location":"abstract s7","hedge_strength":{"p":0.99,"label":"asserted"}},"provenance":{"route":{"via":"TypeSafe API","model":"classifier-1.13.0","endpoint":"https://api.typesafe.ai/v1/systemone","provider":"TypeSafe"},"packing":"one request per paper","requests":1,"input_tokens":7000,"cost_usd_reported":0.0},"compiled_at":"2026-11-04","bank_version":"abstract-p1-20260923","model_version":"classifier-1.13.0","abstract_states":{"claims_first":0.11,"states_limitation":0.57,"human_participants":0.02,"controlled_comparison":0.69,"randomized_assignment":0.01,"states_quantitative_result":0.99,"compares_with_prior_results":0.9,"mentions_code_or_data_availability":0.99},"evidential_basis":{"p":0.9,"value":"observation or experiment","probabilities":{"argument":0.0,"literature":0.0,"proof_or_derivation":0.01,"analysis_of_existing_data":0.0,"observation_or_experiment":0.9,"simulation_or_computation":0.09}},"contribution_type":{"p":1.0,"value":"method or tool","probabilities":{"review":0.0,"method_or_tool":1.0,"empirical_study":0.0,"simulation_study":0.0,"theoretical_result":0.0,"dataset_or_benchmark":0.0,"position_or_commentary":0.0}},"contribution_statement":{"p":1.0,"text":"To overcome this, we introduce Free Draft-and-Verification (FreeDave), a novel training-free and model-free fast decoding algorithm.","location":"abstract s3"}},"coverage":{"tier":"abstract","levels_run":["P abstract"],"not_examined":["full text","figures and tables","references and citing sentences","relations to other papers"],"sources_read":["arXiv listing abstract v4"],"levels_not_run":["S sections","T body sentences","C citations","R cross-paper relations"],"sentences_examined":9,"sentences_in_abstract":9},"measured_accuracy":{"url":"https://pith.science/accuracy","measured_on":"2026-10-02","method":"two independent model families labelled a fresh random sample blind; the figure is how often the published label matches their consensus","sentence_role_agreement":0.92,"main_claim_agreement":0.908,"typed_citation_agreement":0.933,"typed_citation_agreement_95_interval":[0.917,0.945],"review_comments_accurate":0.758,"review_comments_accurate_95_interval":[0.684,0.82],"review_caught_admitted_error_found_or_related":0.448,"review_comments_judged":{"judged_by":["gpt-5.6-luna","grok-4.7"],"judged_against":"Pith's own text copy of each paper: text extracted from the PDF, the first 90,000 characters","measured_on":"2026-10-04","accurate":113,"decided":149,"sampled_reviews":200,"split_by_lane":false,"by_lane":"https://pith.science/accuracy#by-lane"}}}