{ "schema_version": 1, "builder_plan": { "scenario": "You already need task-specific training and want to reduce its memory or compute cost.", "change": "Change only the training technique on a small controlled run after checking prerequisites.", "baseline": "Your current training recipe with the same model, data split, and evaluation.", "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": "Held-out quality, peak memory, training time, and full compute cost to the target quality.", "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": "2609.20807", "title": "Score Centering Stabilizes Off-policy Reinforcement Learning", "summary": "Score centering replaces (or complements) importance-sampling corrections for training-inference mismatch in RL post-training of LLMs by adding an additive drift-canceling term to the score rather than reweighting samples. It costs little compute or memory since the correction is additive, but adds implementation complexity in the RL training loop and requires a reliable estimate of the persistent bias between engines. It can fail if the mismatch is not primarily drift (e.g., high-variance noise rather than persistent bias), if the drift estimate itself is noisy or unstable, or in setups where training and inference engines can simply be kept consistent.", "rationale": "The reader may run RL post-training on a single GPU with quantized inference rollouts, where this low-cost additive correction is worth small-scale validation before replacing importance-sampling machinery.", "technique": "Score centering for RL training-inference mismatch", "family": "treino_eficiente", "recommendation": "testar", "published": "2026-09-17", "as_of": "2026-10-18", "editorial_status": "indexed", "source_url": "https://arxiv.org/abs/2609.20807", "independent_implementations": 1, "total_implementations": 3, "stars": 8022, "citations": 0, "claims": [ { "claim_id": "claim-01", "statement": "Score centering replaces (or complements) importance-sampling corrections for training-inference mismatch in RL post-training of LLMs by adding an additive drift-canceling term to the score rather than reweighting samples. It costs little compute or memory since the correction is additive, but adds implementation complexity in the RL training loop and requires a reliable estimate of the persistent bias between engines. It can fail if the mismatch is not primarily drift (e.g., high-variance noise rather than persistent bias), if the drift estimate itself is noisy or unstable, or in setups where training and inference engines can simply be kept consistent.", "basis": "inferred", "result": "", "baseline": "", "conditions": "", "source_url": "https://arxiv.org/abs/2609.20807", "source_page": null, "source_excerpt": "" }, { "claim_id": "claim-02", "statement": "Score centering alone matches or outperforms importance-sampling-based methods under quantization-induced mismatch, with the gap growing as mismatch worsens; composing it with importance sampling outperforms pure importance-sampling baselines under staleness.", "basis": "inferred", "result": "", "baseline": "", "conditions": "", "source_url": "https://arxiv.org/abs/2609.20807", "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": "" }