{
  "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": "2610.29527",
  "title": "RCSD: Reinforcement and Continual Learning for Domain-Specific Speculative Decoding",
  "summary": "RCSD replaces soft-label distillation of the draft model with a reinforcement-style objective trained on accepted/rejected token pairs collected during speculative inference, optimizing acceptance length directly rather than distribution mimicry. The cost is a custom training pipeline: speculative decoding must already run in production to harvest preference pairs, plus masked-loss training infrastructure and continual updates as traffic shifts, though the resulting draft model remains small and inference-time memory is unchanged. It can fail if production traffic drifts from the training distribution (acceptance rates decay, requiring retraining cycles), if the small team cannot maintain the data-collection loop, or on workloads where speculative decoding itself offers little headroom (e.g., low-batch or highly deterministic generation).",
  "rationale": "The 1.35x end-to-end speedup is meaningful for a 24 GB GPU budget, but the method presupposes an existing speculative-decoding deployment and an ongoing training loop, so it must be validated against the reader's actual draft model and traffic before adoption.",
  "technique": "Reinforcement-trained draft model for speculative decoding",
  "family": "decodificacao_especulativa",
  "recommendation": "testar",
  "published": "2026-10-21",
  "as_of": "2026-10-22",
  "editorial_status": "indexed",
  "source_url": "https://arxiv.org/abs/2610.29527",
  "independent_implementations": 1,
  "total_implementations": 2,
  "stars": 122,
  "citations": null,
  "claims": [
    {
      "claim_id": "claim-01",
      "statement": "RCSD replaces soft-label distillation of the draft model with a reinforcement-style objective trained on accepted/rejected token pairs collected during speculative inference, optimizing acceptance length directly rather than distribution mimicry. The cost is a custom training pipeline: speculative decoding must already run in production to harvest preference pairs, plus masked-loss training infrastructure and continual updates as traffic shifts, though the resulting draft model remains small and inference-time memory is unchanged. It can fail if production traffic drifts from the training distribution (acceptance rates decay, requiring retraining cycles), if the small team cannot maintain the data-collection loop, or on workloads where speculative decoding itself offers little headroom (e.g., low-batch or highly deterministic generation).",
      "basis": "inferred",
      "result": "",
      "baseline": "",
      "conditions": "",
      "source_url": "https://arxiv.org/abs/2610.29527",
      "source_page": null,
      "source_excerpt": ""
    },
    {
      "claim_id": "claim-02",
      "statement": "RCSD improves average acceptance length by 35% over HASS and achieves 1.35x faster end-to-end inference than HASS on practical workloads, without compromising generation quality.",
      "basis": "inferred",
      "result": "",
      "baseline": "",
      "conditions": "",
      "source_url": "https://arxiv.org/abs/2610.29527",
      "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": ""
}