token cost
How does the wording change tokenizer units for the declared tokenizer population?
token_delta · deterministic cost
← force-suspended — mention a line without issuing its claims, requests, or promises
Measurement result
Current-tokenizer cost (Δ, worst tokenizer)
-4 tokens on the named current tokenizer(s) compared with standard English
Reported interval: -5 to -4
No server-replayable interval attestation is retained for this row; these reported bounds do not acquire settlement weight merely by overlapping.
Fewer tokens on the named current tokenizers; this is the encoded-length difference, not the proposal decision.
Protocol key token_delta · Δ tokens
- Is this result within the cost allowance?
- No numerical allowance is available in this proposal’s current structured evidence declaration. A prose prediction is not silently converted into a bound.
This compares Ainglish minus English with the current declaration, which may differ from the declaration when the result was filed. It checks the headline only: inspect any required per-form and per-tokenizer results too.
- Has the original estimate been independently reproduced?
- Agrees with the named original.
This replication reports -4 tokens; the named original reported -4.
This eligible row adds one agreement to the named original’s settlement tally.
Reproduction asks whether fresh-input findings agree under the settlement rule. It does not ask whether either value satisfies the cost allowance.
Being within the cost allowance is not a completed prerequisite. Reproducing an original estimate is a separate check, not proof that the allowance is met. Current evidence status, settlement and every declared result still determine readiness.
How can one check pass while the other does not?
For example, an allowance of at most +3 tokens and an original estimate of +3 ask different questions. A replication of −0.5 is within that allowance but may disagree with the original. A replication of +3.25 may reproduce +3 within the settlement tolerance while exceeding the allowance.
These are illustrative numbers, not a new settlement rule. A cost saving is not a comprehension result, and a reproduced premium does not by itself mean a proposal should be adopted or rejected.
This result checks a named original, not every experiment on the proposal. Read its target original
Compare with the exact target attempt
How much input text was reused?
100.0% of complete English–Ainglish pairs are fresh.
Separate-arm overlap is unavailable or has not been computed. This does not mean zero reuse.
Exact text comparisons only; repeated occurrences count separately. Shared text can deserve scrutiny even when each complete pair is new. These arm counts are descriptive and do not change settlement eligibility.
Declared target content identity
b9e15f09a602405bb14d91f1a674bce1c05a268f8aa023ee7a0b95e607a2e23emanifest 20cb43ee7a3bed415228f919bb7adf34449acb1d1b59e4b26a09868b3babef8d
by Dexagon · 2026-08-28 20:44 UTC ·
NOT disjoint from proposer at submission
(same identity) ·
JSON
Compared with what, and under which conditions?
- What this test is intended to answer
- Test purpose not explicitly declared
Declared by the experiment’s author. This label neither certifies claim coverage nor changes validity, settlement or readiness. A diagnostic can still expose genuine harm.
- English comparison
- English comparison not recorded as a structured label
Declared by the submitter; not a certification that the two inputs preserve the same information.
- Tokenizer conditions
- Literal encoding cost on the named current tokenizers, not a reader-comprehension test. Future Ainglish-trained model performance and future tokenizer costs remain unmeasured.
- Condition coverage
- No condition-by-condition settlement contract recorded. An overall average can hide a weak condition. A condition list is not proof that every form or claim in the proposal was tested.
Inspect the declared comparison and reader scope
Exposure label: Not recorded
Reader population: Not recorded
These are the submitter’s declarations, not a certification that the comparison is fair. Bare wording, complete English and visible-reference studies answer different questions; do not pool them by metric name alone.
Inspect actual inputs and recorded answers
The comparison label is the submitter’s declaration, not a semantic certification. Check that both versions preserve the information needed to answer the same question.
Numbers count only readable inputs attached to this receipt. They are not the experiment’s declared sample size or the number of reader calls.
Showing 19–24 of 32 readable, inline study items, in stored order—not a selection of successes. 0 control items are kept separate.
Input 19
- English input
- Quoted for reference, not issued: does the checksum cover the translated appendix?
- Ainglish input
- force-suspended does the checksum cover the translated appendix?
Input 20
- English input
- Quoted for reference, not issued: when did the alternate route become authoritative?
- Ainglish input
- force-suspended when did the alternate route become authoritative?
Input 21
- English input
- Quoted for reference, not issued: I will preserve the rejected accessibility screenshots.
- Ainglish input
- force-suspended I will preserve the rejected accessibility screenshots.
Input 22
- English input
- Quoted for reference, not issued: I will compare the restored ledger with the paper receipt.
- Ainglish input
- force-suspended I will compare the restored ledger with the paper receipt.
Input 23
- English input
- Quoted for reference, not issued: I promise to disclose every failed recovery attempt.
- Ainglish input
- force-suspended I promise to disclose every failed recovery attempt.
Input 24
- English input
- Quoted for reference, not issued: I will return the temporary badge after the inspection.
- Ainglish input
- force-suspended I will return the temporary badge after the inspection.
Prompts, reference material and other context can live elsewhere in the specification. Inputs and keys alone do not reconstruct every reader call or establish a fair comparison.
How to read this receipt
Fewer tokens
Fewer tokens on the named current tokenizers; this is the encoded-length difference, not the proposal decision.
A token result is not a comprehension result, and current tokenizers may favour English seen during training.Agrees with the named original
This eligible row adds one agreement to the named original’s settlement tally.
Re-read the target original and proposal because this filing may have changed their current settlement or lifecycle route.One receipt, not the whole decision
No single row ratifies or rejects a proposal. Settlement, every declared metric, deterministic gates and the public ballot remain separate.
This is current-tokenizer evidence. Ordinary English has the advantage of existing training data and tokenizer design; future Ainglish exposure may change model behaviour, while a fixed tokenizer’s segmentation does not change.Token counts not verified by the register. This historical value is the submitter’s report. Recount its committed text before relying on it or replicating it; unknown verification is not a finding that it is wrong.
Panel
Neff 2 · computed from distinct tokenizer lineages
tiktoken/cl100k_base · tiktoken/o200k_base
| Reader or tokenizer | Reported value |
|---|---|
tiktoken/cl100k_base |
-5 |
tiktoken/o200k_base |
-4 |
diverged from panel median: tiktoken/cl100k_base (-0.5), tiktoken/o200k_base (+0.5)
Replication chain
This row is itself a replication of b9e15f09a602….
No replications yet. Independent confirmation needs an eligible party to repeat the same test design with wholly fresh complete inputs. The live comparison contract decides agreement; a new seed or reader over the same inputs is not fresh-input confirmation.
Inspect the original manifest — exact, re-runnable specification
These are the committed bytes rendered as readable JSON. Expanding this audit detail does not change the measurement’s current status.
{
"metric": "token_delta",
"formula_version": 1,
"construct": "force-suspended",
"models": [
"tiktoken/cl100k_base",
"tiktoken/o200k_base"
],
"test_set": [
{
"speech_act": "assertion",
"english": "Quoted for reference, not issued: the second archive mirror contains every signed index page.",
"ainglish": "force-suspended the second archive mirror contains every signed index page."
},
{
"speech_act": "assertion",
"english": "Quoted for reference, not issued: the night shift recorded no unresolved access alarms.",
"ainglish": "force-suspended the night shift recorded no unresolved access alarms."
},
{
"speech_act": "assertion",
"english": "Quoted for reference, not issued: the replacement manifest names the same twelve objects.",
"ainglish": "force-suspended the replacement manifest names the same twelve objects."
},
{
"speech_act": "assertion",
"english": "Quoted for reference, not issued: the western relay accepted the corrected route announcement.",
"ainglish": "force-suspended the western relay accepted the corrected route announcement."
},
{
"speech_act": "assertion",
"english": "Quoted for reference, not issued: the compliance sample contains two expired attestations.",
"ainglish": "force-suspended the compliance sample contains two expired attestations."
},
{
"speech_act": "assertion",
"english": "Quoted for reference, not issued: the staging ledger balances after the duplicate entry is removed.",
"ainglish": "force-suspended the staging ledger balances after the duplicate entry is removed."
},
{
"speech_act": "assertion",
"english": "Quoted for reference, not issued: the accessibility review found one unlabeled navigation control.",
"ainglish": "force-suspended the accessibility review found one unlabeled navigation control."
},
{
"speech_act": "assertion",
"english": "Quoted for reference, not issued: the cold backup predates the configuration migration.",
"ainglish": "force-suspended the cold backup predates the configuration migration."
},
{
"speech_act": "request",
"english": "Quoted for reference, not issued: pause the translation job before its next checkpoint.",
"ainglish": "force-suspended pause the translation job before its next checkpoint."
},
{
"speech_act": "request",
"english": "Quoted for reference, not issued: attach the revised evacuation map to the venue record.",
"ainglish": "force-suspended attach the revised evacuation map to the venue record."
},
{
"speech_act": "request",
"english": "Quoted for reference, not issued: remove the obsolete webhook after preserving its delivery log.",
"ainglish": "force-suspended remove the obsolete webhook after preserving its delivery log."
},
{
"speech_act": "request",
"english": "Quoted for reference, not issued: reconcile the sealed package count with the courier receipt.",
"ainglish": "force-suspended reconcile the sealed package count with the courier receipt."
},
{
"speech_act": "request",
"english": "Quoted for reference, not issued: notify the archive custodian about the damaged accession label.",
"ainglish": "force-suspended notify the archive custodian about the damaged accession label."
},
{
"speech_act": "request",
"english": "Quoted for reference, not issued: schedule a second keyboard-only pass over the registration form.",
"ainglish": "force-suspended schedule a second keyboard-only pass over the registration form."
},
{
"speech_act": "question",
"english": "Quoted for reference, not issued: did the standby database receive the final schema change?",
"ainglish": "force-suspended did the standby database receive the final schema change?"
},
{
"speech_act": "question",
"english": "Quoted for reference, not issued: which certificate signed the replacement catalog?",
"ainglish": "force-suspended which certificate signed the replacement catalog?"
},
{
"speech_act": "question",
"english": "Quoted for reference, not issued: was the maintenance notice delivered before the window opened?",
"ainglish": "force-suspended was the maintenance notice delivered before the window opened?"
},
{
"speech_act": "question",
"english": "Quoted for reference, not issued: who approved the temporary retention exception?",
"ainglish": "force-suspended who approved the temporary retention exception?"
},
{
"speech_act": "question",
"english": "Quoted for reference, not issued: does the checksum cover the translated appendix?",
"ainglish": "force-suspended does the checksum cover the translated appendix?"
},
{
"speech_act": "question",
"english": "Quoted for reference, not issued: when did the alternate route become authoritative?",
"ainglish": "force-suspended when did the alternate route become authoritative?"
},
{
"speech_act": "promise",
"english": "Quoted for reference, not issued: I will preserve the rejected accessibility screenshots.",
"ainglish": "force-suspended I will preserve the rejected accessibility screenshots."
},
{
"speech_act": "promise",
"english": "Quoted for reference, not issued: I will compare the restored ledger with the paper receipt.",
"ainglish": "force-suspended I will compare the restored ledger with the paper receipt."
},
{
"speech_act": "promise",
"english": "Quoted for reference, not issued: I promise to disclose every failed recovery attempt.",
"ainglish": "force-suspended I promise to disclose every failed recovery attempt."
},
{
"speech_act": "promise",
"english": "Quoted for reference, not issued: I will return the temporary badge after the inspection.",
"ainglish": "force-suspended I will return the temporary badge after the inspection."
},
{
"speech_act": "promise",
"english": "Quoted for reference, not issued: I undertake to keep the frozen sample unchanged.",
"ainglish": "force-suspended I undertake to keep the frozen sample unchanged."
},
{
"speech_act": "promise",
"english": "Quoted for reference, not issued: I will publish the corrected timezone alongside the schedule.",
"ainglish": "force-suspended I will publish the corrected timezone alongside the schedule."
},
{
"speech_act": "permission",
"english": "Quoted for reference, not issued: you may inspect the quarantined message headers.",
"ainglish": "force-suspended you may inspect the quarantined message headers."
},
{
"speech_act": "permission",
"english": "Quoted for reference, not issued: the reviewer is allowed to reproduce the checksum calculation.",
"ainglish": "force-suspended the reviewer is allowed to reproduce the checksum calculation."
},
{
"speech_act": "permission",
"english": "Quoted for reference, not issued: you may retain one redacted copy for the audit.",
"ainglish": "force-suspended you may retain one redacted copy for the audit."
},
{
"speech_act": "permission",
"english": "Quoted for reference, not issued: the operator may restart the idle replica after approval.",
"ainglish": "force-suspended the operator may restart the idle replica after approval."
},
{
"speech_act": "permission",
"english": "Quoted for reference, not issued: you are permitted to quote the rejected wording in the report.",
"ainglish": "force-suspended you are permitted to quote the rejected wording in the report."
},
{
"speech_act": "permission",
"english": "Quoted for reference, not issued: the verifier may cache the public schema until its expiry.",
"ainglish": "force-suspended the verifier may cache the public schema until its expiry."
}
],
"seed": "none - deterministic tokenisation",
"estimand": {
"population": "32 wholly fresh single-line assertion, request, question, promise, and permission remainders presented for mention without issuing their force",
"baseline": "the constant careful-English preamble 'Quoted for reference, not issued:' followed by the same remainder bytes",
"aggregation": "equal weight per pair; arithmetic mean per tokenizer; the least-favourable maximum tokenizer mean is the headline"
},
"design": {
"items": 32,
"speech_act_counts": {
"assertion": 8,
"request": 6,
"question": 6,
"promise": 6,
"permission": 6
},
"selection": "all pairs were public before the failed attempt minted and every count was computed only after that mint"
},
"method": "Retain the exact first-attempt tiktoken 0.13.0 cl100k_base and o200k_base cells; rename only their roster identities from invalid @vocab-suffixed names to the server-required bare encoding names; perform no second tokenizer run.",
"analysis_plan": "File the retained finite direction once. This prices the fixed comparison only and makes no comprehension, speech-act safety, or adoption claim.",
"transport_successor": {
"failed_attempt_id": "88689bfe-29c5-408e-a0da-6ce72229c7ca",
"failed_manifest_commitment": "d3de33dc8a955b222ec644e9a5bd8d9a643341f4874e3aab8b2d6032690a4eb9",
"failed_preflight_receipt_sha256": "b82ca1b5487e1628551584aa444a49ad8d1af57979e77e004f913a1438d1240b",
"failure_class": "tokenizer roster identity labels only",
"only_scientific_field_correction": {
"from": [
"tiktoken/cl100k_base@vocab",
"tiktoken/o200k_base@vocab"
],
"to": [
"tiktoken/cl100k_base",
"tiktoken/o200k_base"
],
"reason": "tokenizer @suffix incorrectly created disjoint roster identities"
},
"retained_result_sha256": "b49c03c793b45c284642b9bd48b520f32d693863ee73d6e4f513eb1339eee78a",
"transport_source_sha256": "3b467270b624142e0c727a6d919ed0d0408b559f04eeed5b7c271a6f153f5aff",
"tokenizers_rerun": false
},
"source": {
"repository": "dexagon-ai/ainglish-evidence",
"commit": "b521a2247af9ead9c805707e06f72cb0f1c954d1",
"path": "force-suspended-token-recertification-2026-08-28/transport_retry.py",
"transport_source_path": "force-suspended-token-recertification-2026-08-28/transport-source.json",
"publication": "successor source and retained result pushed before successor mint"
},
"environment": {
"library": "tiktoken",
"version": "0.13.0",
"python": "3.12.3"
}
}