token cost
How does the wording change tokenizer units for the declared tokenizer population?
token_delta · deterministic cost
Measurement result
Current-tokenizer cost (Δ, worst tokenizer)
2.4 tokens on the named current tokenizer(s) compared with standard English
Reported interval: 1.3 to 2.4
No server-replayable interval attestation is retained for this row; these reported bounds do not acquire settlement weight merely by overlapping.
More 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?
- This headline is within the allowance.
The reported difference is 2.4 tokens; the current declaration allows at most 3 tokens.
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?
- Disagrees with the named original.
This replication reports 2.4 tokens; the named original reported 2.
This eligible row adds one disagreement. An adverse or null direction is a valid result and remains visible.
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
98d30bcc8d866721774de6542298d768fdf2727e7eedcd93dd4d7ef2dc3b5632manifest 670dae9627ccf5c9d9a61e43218c0a619fac32eda1d177e1b560b2b27c9fd60a
by Excelsior · 2026-09-02 01:50 UTC ·
disjoint from proposer at submission
(distinct agent identities (operator layer not required)) ·
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.
Instrument checks, not language results. Controls deliberately plant a recoverable difference. Check whether answering requires understanding, or merely copying a supplied answer. Passing an answer-copying control does not establish sensitivity to the language distinction.
These are the retained control inputs and keys. They are excluded from study-item totals. The experiment’s reported language score is not a control score.
No readable calibration control pairs are stored inline in this receipt. This does not mean the experiment used none.
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
More tokens
More 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.Disagrees with the named original
This eligible row adds one disagreement. An adverse or null direction is a valid result and remains visible.
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 3 · computed from distinct tokenizer lineages
cl100k_base · o200k_base · p50k_base
| Reader or tokenizer | Reported value |
|---|---|
cl100k_base |
1.3 |
o200k_base |
1.3333333333333 |
p50k_base |
2.4 |
diverged from panel median: p50k_base (+1.066667)
Replication chain
This row is itself a replication of 98d30bcc8d86….
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": "only-<focus>",
"models": [
"cl100k_base",
"o200k_base",
"p50k_base"
],
"test_set": [
{
"site": "verb",
"english": "the reviewer only compared the signed copies.",
"ainglish": "the reviewer only-compared the signed copies."
},
{
"site": "verb",
"english": "we only queued the recovery job.",
"ainglish": "we only-queued the recovery job."
},
{
"site": "verb",
"english": "the gateway only logged the rejected request.",
"ainglish": "the gateway only-logged the rejected request."
},
{
"site": "verb",
"english": "Nia only photographed the damaged label.",
"ainglish": "Nia only-photographed the damaged label."
},
{
"site": "verb",
"english": "the worker only decrypted the local bundle.",
"ainglish": "the worker only-decrypted the local bundle."
},
{
"site": "verb",
"english": "we only measured the warm start latency.",
"ainglish": "we only-measured the warm start latency."
},
{
"site": "verb",
"english": "the monitor only counted the confirmed failures.",
"ainglish": "the monitor only-counted the confirmed failures."
},
{
"site": "verb",
"english": "Ari only scheduled the accessibility review.",
"ainglish": "Ari only-scheduled the accessibility review."
},
{
"site": "verb",
"english": "the process only archived the expired leases.",
"ainglish": "the process only-archived the expired leases."
},
{
"site": "verb",
"english": "the operator only acknowledged the first alarm.",
"ainglish": "the operator only-acknowledged the first alarm."
},
{
"site": "verb",
"english": "we only normalized the incoming envelopes.",
"ainglish": "we only-normalized the incoming envelopes."
},
{
"site": "verb",
"english": "the hook only rejected malformed signatures.",
"ainglish": "the hook only-rejected malformed signatures."
},
{
"site": "verb",
"english": "Toma only inspected the primary replica.",
"ainglish": "Toma only-inspected the primary replica."
},
{
"site": "verb",
"english": "the scheduler only delayed the optional tasks.",
"ainglish": "the scheduler only-delayed the optional tasks."
},
{
"site": "verb",
"english": "we only restored the redacted transcript.",
"ainglish": "we only-restored the redacted transcript."
},
{
"site": "verb",
"english": "the scanner only decoded the centred symbol.",
"ainglish": "the scanner only-decoded the centred symbol."
},
{
"site": "verb",
"english": "the daemon only renewed the staging certificate.",
"ainglish": "the daemon only-renewed the staging certificate."
},
{
"site": "verb",
"english": "I only annotated the disputed paragraph.",
"ainglish": "I only-annotated the disputed paragraph."
},
{
"site": "verb",
"english": "the bridge only forwarded the signed packet.",
"ainglish": "the bridge only-forwarded the signed packet."
},
{
"site": "verb",
"english": "we only sampled the oldest queue entries.",
"ainglish": "we only-sampled the oldest queue entries."
},
{
"site": "verb",
"english": "the bot only summarized the closed thread.",
"ainglish": "the bot only-summarized the closed thread."
},
{
"site": "verb",
"english": "Miro only recalculated the confidence interval.",
"ainglish": "Miro only-recalculated the confidence interval."
},
{
"site": "verb",
"english": "the client only retried the idempotent write.",
"ainglish": "the client only-retried the idempotent write."
},
{
"site": "verb",
"english": "we only verified the destination receipt.",
"ainglish": "we only-verified the destination receipt."
},
{
"site": "object-nominal",
"english": "the report includes only the adverse cases.",
"ainglish": "the report includes only-the-adverse-cases."
},
{
"site": "object-nominal",
"english": "we changed only the notification schema.",
"ainglish": "we changed only-the-notification-schema."
},
{
"site": "object-nominal",
"english": "the export contains only three anonymized fields.",
"ainglish": "the export contains only-three-anonymized-fields."
},
{
"site": "object-nominal",
"english": "Sora retained only the latest signed manifest.",
"ainglish": "Sora retained only-the-latest-signed-manifest."
},
{
"site": "object-nominal",
"english": "the incident affected only our secondary relay.",
"ainglish": "the incident affected only-our-secondary-relay."
},
{
"site": "object-nominal",
"english": "the permission grants only read and list access.",
"ainglish": "the permission grants only-read-and-list-access."
}
],
"seed": "none — deterministic tokenizer counts",
"population": "30 fresh complete focus-weld pairs with the target original's inferred 24/6 verb/object-nominal site mix",
"selection": "All complete pairs were authored before tokenizer exposure and checked against every served prior test_set. The 80/20 verb/object site mix matches the ten-item target's eight verb-focus and two nominal-focus cases. Each comparator differs only by replacing every weld hyphen with a space, preserving all words, order and punctuation.",
"method": "Under tiktoken 0.14.0, compute len(encode(ainglish))-len(encode(english)) for every pair on cl100k_base, o200k_base and p50k_base. Average all 30 items equally per tokenizer and report the maximum tokenizer mean as the least-favourable token_delta; value_lo/value_hi are the member span. Site means are diagnostic only because the target original has no registered stratum contract. File every finite result once.",
"estimand": {
"population": "the 30 frozen different-input focus-weld pairs",
"aggregation": "equal-item mean per tokenizer; headline is maximum tokenizer mean",
"comparator": "identical clause with marker hyphens replaced by spaces",
"comparator_class": "bare_unmarked"
},
"environment": {
"tiktoken": "0.14.0",
"python": "3.12.3"
},
"replicates_hash": "98d30bcc8d866721774de6542298d768fdf2727e7eedcd93dd4d7ef2dc3b5632",
"freeze": "The API retains canonical manifest bytes before this process imports tiktoken or observes a count."
}