{ "schema_version": 1, "builder_plan": { "scenario": "Your self-hosted model spends too long generating answers token by token.", "change": "Check whether a compatible draft model and serving integration exist before renting compute.", "baseline": "The same target with speculation off, then your existing drafter if you have one.", "steps": [ "Locate the paper's implementation and check its license, model support, data needs, and hardware requirements. Stop here if you cannot obtain the required artifacts.", "Save a small set of representative product tasks and expected outcomes. Keep evaluation tasks out of any training or tuning, and choose quality and cost limits before running.", "Run the baseline and the proposed change on identical inputs. Keep other settings fixed and repeat paired runs to expose variation.", "Save per-task outputs, failures, timings, and all usage costs. Review regressions before expanding the test." ], "measure": "Accepted draft length, tokens per second, tail latency, output quality, and total GPU cost.", "decision": "Keep the change only if it meets your preselected quality and latency limits and improves the metric you care about after setup and operating costs. Otherwise retain the baseline. This is an exploration plan, not a validated recommendation." }, "builder_review": null, "arxiv_id": "2609.39972", "title": "UBTree: Parallel Tree Drafting via Unigram and Bigram Models for Speculative Decoding", "summary": "UBTree replaces single-path autoregressive drafters (and prior parallel drafters such as DARTree) with a tree drafter in which a unigram proposer generates per-position candidates and a bigram selector, trained with renormalized KL on high-temperature data, scores transitions between adjacent candidates. The cost is training and serving two additional lightweight draft components, extra tree-construction and verification logic in the inference path, and the overhead of candidate scoring per step. The approach can fail on high-entropy workloads if the tree still lacks sufficient branch diversity, if the proposer-selector training distribution diverges from production traffic, or if per-step drafting overhead erodes the speedup at small batch sizes typical of a single 24 GB GPU.", "rationale": "The claimed speedup is substantial and the drafter components are lightweight enough for a single-GPU setup, but the gains depend on retraining the proposer and selector on data matching the reader's target model and workload, so adoption requires workload-specific validation rather than immediate deployment.", "technique": "Unigram-bigram tree drafting for speculative decoding", "family": "decodificacao_especulativa", "recommendation": "testar", "published": "2026-09-30", "as_of": "2026-10-18", "editorial_status": "indexed", "source_url": "https://arxiv.org/abs/2609.39972", "independent_implementations": 0, "total_implementations": 1, "stars": 121, "citations": null, "claims": [ { "claim_id": "claim-01", "statement": "UBTree replaces single-path autoregressive drafters (and prior parallel drafters such as DARTree) with a tree drafter in which a unigram proposer generates per-position candidates and a bigram selector, trained with renormalized KL on high-temperature data, scores transitions between adjacent candidates. The cost is training and serving two additional lightweight draft components, extra tree-construction and verification logic in the inference path, and the overhead of candidate scoring per step. The approach can fail on high-entropy workloads if the tree still lacks sufficient branch diversity, if the proposer-selector training distribution diverges from production traffic, or if per-step drafting overhead erodes the speedup at small batch sizes typical of a single 24 GB GPU.", "basis": "inferred", "result": "", "baseline": "", "conditions": "", "source_url": "https://arxiv.org/abs/2609.39972", "source_page": null, "source_excerpt": "" }, { "claim_id": "claim-02", "statement": "Average speedup of 5.84-6.94x over autoregressive decoding across seven benchmarks with Qwen3-4B and Qwen3-8B, outperforming DARTree in all 28 comparisons.", "basis": "inferred", "result": "", "baseline": "", "conditions": "", "source_url": "https://arxiv.org/abs/2609.39972", "source_page": null, "source_excerpt": "" } ], "exposure_map": [ { "dimension": "quality", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "compute", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "latency", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "operations", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "compatibility", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "security", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "data_and_training", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "reproducibility", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" } ], "risks": [], "minimum_test": [], "open_questions": [], "independent_tests": [], "lab_reproductions": [], "report_available": false, "validation_tier": "unknown", "evidence_tier": "unknown", "infrastructure_basis": "unknown", "equations": [], "equations_status": "not_fetched", "core_kind": "", "equations_fetched_at": "", "equations_source_url": "", "equations_source_sha256": "", "exposures_reviewed_at": "", "exposures_source_sha256": "" }