{ "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": "2610.29545", "title": "PACE: Periodic and Channel-Aligned Encoding of Four-Bit Recurrent States", "summary": "PACE replaces per-token rounding of the recurrent state in linear-attention and hybrid models (e.g., 8-bit DAMP or Q-Mamba) with post-hoc 4-bit quantization that rounds only once every few tokens, shares scales per column, and keeps error-sensitive heads at 8 bits. It requires a small token buffer between rounding points and mixed-precision head handling, adding modest kernel and serving complexity; accuracy on the 4B model is not fully preserved at 4 bits. Failure modes include quality degradation on smaller models, dependence on chunked-kernel serving stacks that support buffered decoding, and sensitivity to workloads dominated by long concurrent generations where state memory is not the actual bottleneck.", "rationale": "It directly increases concurrent sequence capacity on a single 24GB GPU for hybrid linear-attention models, but gains depend on the reader using such architectures and a compatible chunked inference stack, so workload-specific validation is needed before adoption.", "technique": "Post-hoc 4-bit recurrent state quantization", "family": "quantizacao", "recommendation": "testar", "published": "2026-10-21", "as_of": "2026-10-22", "editorial_status": "indexed", "source_url": "https://arxiv.org/abs/2610.29545", "independent_implementations": 1, "total_implementations": 2, "stars": 122, "citations": null, "claims": [ { "claim_id": "claim-01", "statement": "PACE replaces per-token rounding of the recurrent state in linear-attention and hybrid models (e.g., 8-bit DAMP or Q-Mamba) with post-hoc 4-bit quantization that rounds only once every few tokens, shares scales per column, and keeps error-sensitive heads at 8 bits. It requires a small token buffer between rounding points and mixed-precision head handling, adding modest kernel and serving complexity; accuracy on the 4B model is not fully preserved at 4 bits. Failure modes include quality degradation on smaller models, dependence on chunked-kernel serving stacks that support buffered decoding, and sensitivity to workloads dominated by long concurrent generations where state memory is not the actual bottleneck.", "basis": "inferred", "result": "", "baseline": "", "conditions": "", "source_url": "https://arxiv.org/abs/2610.29545", "source_page": null, "source_excerpt": "" }, { "claim_id": "claim-02", "statement": "On three hybrid models from 8B to 35B, PACE with mixed 4/8-bit heads matches 8-bit DAMP quality at about two thirds of its state memory, enabling more concurrent sequences per accelerator.", "basis": "inferred", "result": "", "baseline": "", "conditions": "", "source_url": "https://arxiv.org/abs/2610.29545", "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": "" }