{ "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": "2610.19801", "title": "Delta Tokenizer: Reference-Identical Tokenization of Session Deltas and Segment-Parallelism for Long Prompts on Multicore CPUs", "summary": "Replaces full re-tokenization of each request with reuse of the prior session's token sequence, re-encoding only the appended text and a boundary tail, with an exact parallel chunking path for cold long prompts. Cost is additional CPU-side state per session (retained token sequences), integration into the serving front end (currently demonstrated in SGLang), and a shared CPU pool scheduler that adds operational complexity. It can fail if sessions are short-lived or prompts change mid-conversation rather than by pure append, if CPU cores are already saturated by the tokenizer pool, or if tokenizer versions diverge between stored state and the serving engine.", "rationale": "It is a CPU-only, reference-exact drop-in for the serving front end that directly reduces TTFT in multi-turn agent workloads, requiring no GPU, no model changes, and no training budget.", "technique": "Incremental delta tokenization for session prompts", "family": "serving_e_batching", "recommendation": "adotar", "published": "2026-10-14", "as_of": "2026-10-18", "editorial_status": "indexed", "source_url": "https://arxiv.org/abs/2610.19801", "independent_implementations": 0, "total_implementations": 1, "stars": 6, "citations": 0, "claims": [ { "claim_id": "claim-01", "statement": "Replaces full re-tokenization of each request with reuse of the prior session's token sequence, re-encoding only the appended text and a boundary tail, with an exact parallel chunking path for cold long prompts. Cost is additional CPU-side state per session (retained token sequences), integration into the serving front end (currently demonstrated in SGLang), and a shared CPU pool scheduler that adds operational complexity. It can fail if sessions are short-lived or prompts change mid-conversation rather than by pure append, if CPU cores are already saturated by the tokenizer pool, or if tokenizer versions diverge between stored state and the serving engine.", "basis": "inferred", "result": "", "baseline": "", "conditions": "", "source_url": "https://arxiv.org/abs/2610.19801", "source_page": null, "source_excerpt": "" }, { "claim_id": "claim-02", "statement": "Tokenization of a session turn is roughly 40x faster than the Hugging Face reference tokenizer; inside SGLang, median TTFT drops 59% at 100K tokens and P99 TTFT for agent sessions falls 9.5x under load and up to 23x when long requests occupy the front end.", "basis": "inferred", "result": "", "baseline": "", "conditions": "", "source_url": "https://arxiv.org/abs/2610.19801", "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": "" }