{
  "schema_version": 1,
  "builder_plan": {
    "scenario": "Your inference service slows down when several customers send requests together.",
    "change": "Replay a recorded request trace with only the proposed serving policy changed.",
    "baseline": "Your current scheduler on the same model and hardware.",
    "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": "Requests meeting your latency target, throughput, failures, and cost per successful request.",
    "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": "2609.38981",
  "title": "Vosti: Specifying, Implementing, and Verifying Deterministic LLM Inference",
  "summary": "Vosti replaces the ad hoc deterministic modes of engines like vLLM and SGLang, which the authors show can still produce divergent logits under batching, chunking, and KV-cache variations, with an engine whose scheduler and paged prefix-sharing KV-cache are formally verified against a bitwise-determinism specification. The cost is adopting a new, research-stage inference engine: kernel selection is fixed independently of runtime state, which constrains scheduling flexibility, and performance is only shown to be comparable to vLLM's batch-invariant mode on decode-heavy workloads rather than superior. What can fail is practical: an immature codebase with unknown model and hardware coverage, a Triton/Verus proof boundary that may not cover kernels outside the analyzed set, and operational risk if the engine lacks features (LoRA, quantization, wide model support) the reader's deployment requires.",
  "rationale": "Bitwise-deterministic inference is genuinely valuable for reproducibility, debugging, and regression testing in production, but a formally verified research engine carries adoption risk that a small team should validate on its own workloads before replacing a maintained engine like vLLM or SGLang.",
  "technique": "Formally verified deterministic inference engine",
  "family": "serving_e_batching",
  "recommendation": "testar",
  "published": "2026-09-30",
  "as_of": "2026-10-18",
  "editorial_status": "indexed",
  "source_url": "https://arxiv.org/abs/2609.38981",
  "independent_implementations": 0,
  "total_implementations": 0,
  "stars": 0,
  "citations": 0,
  "claims": [
    {
      "claim_id": "claim-01",
      "statement": "Vosti replaces the ad hoc deterministic modes of engines like vLLM and SGLang, which the authors show can still produce divergent logits under batching, chunking, and KV-cache variations, with an engine whose scheduler and paged prefix-sharing KV-cache are formally verified against a bitwise-determinism specification. The cost is adopting a new, research-stage inference engine: kernel selection is fixed independently of runtime state, which constrains scheduling flexibility, and performance is only shown to be comparable to vLLM's batch-invariant mode on decode-heavy workloads rather than superior. What can fail is practical: an immature codebase with unknown model and hardware coverage, a Triton/Verus proof boundary that may not cover kernels outside the analyzed set, and operational risk if the engine lacks features (LoRA, quantization, wide model support) the reader's deployment requires.",
      "basis": "inferred",
      "result": "",
      "baseline": "",
      "conditions": "",
      "source_url": "https://arxiv.org/abs/2609.38981",
      "source_page": null,
      "source_excerpt": ""
    },
    {
      "claim_id": "claim-02",
      "statement": "Produces bitwise-identical logits across every tested execution variation, with performance comparable to vLLM's batch-invariant mode on decode-heavy workloads; no speedup factor is claimed.",
      "basis": "inferred",
      "result": "",
      "baseline": "",
      "conditions": "",
      "source_url": "https://arxiv.org/abs/2609.38981",
      "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": ""
}