{ "schema_version": 1, "builder_plan": { "scenario": "A self-hosted assistant is too expensive because its model uses most of the GPU memory.", "change": "Try the paper's supported lower-precision weights on a copy of your serving setup.", "baseline": "The same model at its original precision, on the same GPU.", "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": "Peak memory, task pass rate, latency, and total serving cost at the same context and load.", "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": "2609.21450", "title": "Understanding LLM Quantization through Activation-Guided Compensation and Orthogonal Residuals", "summary": "The method replaces gradient-trained rotation and scaling pipelines such as SpinQuant with a closed-form, activation-guided error decomposition that selects Hadamard rotations, sign patterns, and L2/L-infinity channel scaling without backpropagation. It costs only calibration-forward passes to compute statistics and sample sign patterns, adding negligible engineering complexity relative to standard PTQ toolchains, with no runtime overhead since transformations fold into weights. It can fail when persistent outlier structure violates the residual-bound assumptions or when calibration data is unrepresentative, in which case aggressive W4A4 accuracy still degrades relative to wider-bit schemes.", "rationale": "A gradient-free PTQ method matching SpinQuant fits a single 24 GB GPU workflow and low serving memory budgets, but the abstract reports parity rather than measured gains, so workload-specific validation on the reader's models and data is required before production adoption.", "technique": "Activation-guided error decomposition for W4A4 quantization", "family": "quantizacao", "recommendation": "testar", "published": "2026-09-18", "as_of": "2026-10-18", "editorial_status": "indexed", "source_url": "https://arxiv.org/abs/2609.21450", "independent_implementations": 0, "total_implementations": 0, "stars": 0, "citations": 0, "claims": [ { "claim_id": "claim-01", "statement": "The method replaces gradient-trained rotation and scaling pipelines such as SpinQuant with a closed-form, activation-guided error decomposition that selects Hadamard rotations, sign patterns, and L2/L-infinity channel scaling without backpropagation. It costs only calibration-forward passes to compute statistics and sample sign patterns, adding negligible engineering complexity relative to standard PTQ toolchains, with no runtime overhead since transformations fold into weights. It can fail when persistent outlier structure violates the residual-bound assumptions or when calibration data is unrepresentative, in which case aggressive W4A4 accuracy still degrades relative to wider-bit schemes.", "basis": "inferred", "result": "", "baseline": "", "conditions": "", "source_url": "https://arxiv.org/abs/2609.21450", "source_page": null, "source_excerpt": "" }, { "claim_id": "claim-02", "statement": "Backpropagation-free W4A4 configurations achieve performance competitive with gradient-trained SpinQuant across eight Llama and Mistral models; no numeric factor reported.", "basis": "inferred", "result": "", "baseline": "", "conditions": "", "source_url": "https://arxiv.org/abs/2609.21450", "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": "" }