These are literal recorded values, not a compatibility test. Matching declarations do not prove equivalent inputs or fair scoring. A missing structured field may be described elsewhere in the immutable specification.
What was measured · Same recorded value
First result
token_delta
Second result
token_delta
English comparison declarations · Not recorded on either side
First result
Not recorded in this structured field
Second result
Not recorded in this structured field
Tested population · Same recorded value
First result
24 frozen complete operational claims across 24 domains, each with a claim-attached observable falsifier
Second result
24 frozen complete operational claims across 24 domains, each with a claim-attached observable falsifier
Unit tested · Same recorded value
First result
one complete operational claim with its falsifier
Second result
one complete operational claim with its falsifier
Named instruments (including order) · Same recorded value
First result
[
"cl100k_base",
"o200k_base",
"p50k_base"
]
Second result
[
"cl100k_base",
"o200k_base",
"p50k_base"
]
Reader exposure · Not recorded on either side
First result
Not recorded in this structured field
Second result
Not recorded in this structured field
Exposure window · Not recorded on either side
First result
Not recorded in this structured field
Second result
Not recorded in this structured field
How results combine · Same recorded value
First result
{
"reducer": "least_favourable",
"rule": "equal item mean per tokenizer, then least-favourable maximum tokenizer mean; retain the single unless stratum"
}
Second result
{
"reducer": "least_favourable",
"rule": "equal item mean per tokenizer, then least-favourable maximum tokenizer mean; retain the single unless stratum"
}
Scoring declaration · Not recorded on either side
First result
Not recorded in this structured field
Second result
Not recorded in this structured field
Conditions (including order) · Same recorded value
First result
[
{
"id": "unless",
"weight": 1
}
]
Second result
[
{
"id": "unless",
"weight": 1
}
]
First result: full recordtoken cost · -2.625 tokens per declared item
Counts in current evidence decisions. This row currently contributes to evidence decisions. Its direction is separate from whether the proposal is ready for adoption.
What was measured
token cost · token_delta How does the wording change tokenizer units for the declared tokenizer population?
Reported result
-2.625 tokens per declared item Reported interval: -3.7083333333333 to -2.625.
Fewer tokens on the named current tokenizers; this is the encoded-length difference, not the proposal decision.
English comparison
English comparison not recorded as a structured label
Declared by the submitter; not a certification of equivalent information.
Tokenizer conditions
Literal encoding cost on the named current tokenizers; not comprehension.
Named instruments
cl100k_base, o200k_base, p50k_base
Reader population not separately declared.
Conditions covered
Separate outcomes retained for all 1 declared conditions
unless
An overall average can hide a weak condition. A condition list is not proof that every form or claim in the proposal was tested.
Settlement role
Confirmed by eligible settlement
Eligible fresh-input replications currently give this original a settlement majority.
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?
Confirmed by eligible settlement.
Eligible fresh-input replications currently give this original a settlement majority.
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.
Does the overall result hide differences between conditions?
Every stored condition, without new pooling. Differences and intervals use tokens. Condition names come from the frozen experiment.
Condition
Reported difference
Reported interval
unless
-2.625
Not recorded
A missing condition interval is not zero uncertainty. An overall interval cannot substitute for agreement in every load-bearing condition.
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 1–6 of 24 readable, inline study items, in stored order—not a selection of successes. 0 control items are kept separate.
Each result has its own input pages. Positions across the two studies do not imply matched cases.
Input 1 · vaccine-fridges
English input
The temperature log covers every vaccine fridge — that claim fails if sensor 14 was offline during the night shift.
Ainglish input
The temperature log covers every vaccine fridge unless(sensor 14 was offline during the night shift).
Condition
unless
Input 2 · berth-plan
English input
The berth allocation contains no vessel conflict — that claim fails if tug arrival B17 was entered in local time.
Ainglish input
The berth allocation contains no vessel conflict unless(tug arrival B17 was entered in local time).
Condition
unless
Input 3 · quarterly-ledger
English input
The quarterly ledger balances — that claim fails if a foreign-currency adjustment is still pending.
Ainglish input
The quarterly ledger balances unless(a foreign-currency adjustment is still pending).
Condition
unless
Input 4 · wildlife-count
English input
The wildlife count includes every tagged otter — that claim fails if receiver C lost packets near the estuary.
Ainglish input
The wildlife count includes every tagged otter unless(receiver C lost packets near the estuary).
Condition
unless
Input 5 · build-attestation
English input
The build attestation identifies every dependency — that claim fails if the plugin resolver used an unrecorded cache.
Ainglish input
The build attestation identifies every dependency unless(the plugin resolver used an unrecorded cache).
Condition
unless
Input 6 · appeal-register
English input
The appeal register is complete — that claim fails if a sealed filing was entered after the export.
Ainglish input
The appeal register is complete unless(a sealed filing was entered after the export).
Condition
unless
Recorded input digest: 9dc876e1bbb6a51281138cfcbe7336b81afa940b980eff05a75a0423d544e59c
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.
Declared population, method and retained outcomes
Compared with
unless(<F>) versus the complete registered disclosure 'that claim fails if F' with identical claim and falsifier
Tested population
24 frozen complete operational claims across 24 domains, each with a claim-attached observable falsifier
Unit tested
one complete operational claim with its falsifier
How results combine
equal item mean per tokenizer, then least-favourable maximum tokenizer mean; retain the single unless stratum
These are the study author’s declarations. A finding applies to this tested scope; this summary does not establish that another study is comparable.
Absolute arm results, reader-specific results and condition results below are retained values, not a newly pooled analysis. Accuracy arms use fractions from 0 to 1; their difference uses percentage points.
Counts in current evidence decisions. This row currently contributes to evidence decisions. Its direction is separate from whether the proposal is ready for adoption.
What was measured
token cost · token_delta How does the wording change tokenizer units for the declared tokenizer population?
Reported result
-2.5 tokens per declared item Reported interval: -3.2083333333333 to -2.5.
Fewer tokens on the named current tokenizers; this is the encoded-length difference, not the proposal decision.
English comparison
English comparison not recorded as a structured label
Declared by the submitter; not a certification of equivalent information.
Tokenizer conditions
Literal encoding cost on the named current tokenizers; not comprehension.
Named instruments
cl100k_base, o200k_base, p50k_base
Reader population not separately declared.
Conditions covered
Separate outcomes retained for all 1 declared conditions
unless
An overall average can hide a weak condition. A condition list is not proof that every form or claim in the proposal was tested.
Settlement role
Agrees with the named original
This eligible row adds one agreement to the named original’s settlement tally.
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 -2.5 tokens; the named original reported -2.625.
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
Every declared condition must agree. Overlapping overall intervals alone do not confirm this original.
How much input text was reused?
100.0% of complete English–Ainglish pairs are fresh.
0 of 24 English inputs reuse text from either side of the original.
0 of 24 Ainglish inputs reuse text from either side of the original.
Declared item-bank digests: different. This compares bank identity, not shared sentences; different bank digests can still contain identical pairs.
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.
Does the overall result hide differences between conditions?
Every stored condition, without new pooling. Differences and intervals use tokens. Condition names come from the frozen experiment.
Condition
Reported difference
Reported interval
unless
-2.5
Not recorded
A missing condition interval is not zero uncertainty. An overall interval cannot substitute for agreement in every load-bearing condition.
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 1–6 of 24 readable, inline study items, in stored order—not a selection of successes. 0 control items are kept separate.
Each result has its own input pages. Positions across the two studies do not imply matched cases.
Input 1 · ex-unless-20260919-sample-band
English input
Every monitored sample remained between two and eight degrees Celsius — that claim fails if a recorded reading for sample Sable is nine degrees Celsius.
Ainglish input
Every monitored sample remained between two and eight degrees Celsius unless(a recorded reading for sample Sable is nine degrees Celsius).
Condition
unless
Input 2 · ex-unless-20260919-pilot-assignment
English input
Every departure in the port log has a pilot assigned — that claim fails if departure Kestrel has no assigned pilot in the log.
Ainglish input
Every departure in the port log has a pilot assigned unless(departure Kestrel has no assigned pilot in the log).
Condition
unless
Input 3 · ex-unless-20260919-invoice-identity
English input
The invoice register uses each invoice number once — that claim fails if invoice 731 appears in two separate register entries.
Ainglish input
The invoice register uses each invoice number once unless(invoice 731 appears in two separate register entries).
Condition
unless
Input 4 · ex-unless-20260919-nest-location
English input
Every surveyed nest has a recorded location — that claim fails if surveyed nest Clover has an empty location field.
Ainglish input
Every surveyed nest has a recorded location unless(surveyed nest Clover has an empty location field).
Condition
unless
Input 5 · ex-unless-20260919-release-checksums
English input
Every executable in the release has a listed checksum — that claim fails if released executable Lantern has no checksum entry.
Ainglish input
Every executable in the release has a listed checksum unless(released executable Lantern has no checksum entry).
Condition
unless
Input 6 · ex-unless-20260919-judgment-paragraphs
English input
The published judgment includes all numbered paragraphs — that claim fails if numbered paragraph 47 is absent from the published judgment.
Ainglish input
The published judgment includes all numbered paragraphs unless(numbered paragraph 47 is absent from the published judgment).
Condition
unless
Recorded input digest: 28601dc415764ff4c3d74e1047598d257cef07f36edf9ddfe73175739a1cd9e3
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.
Declared population, method and retained outcomes
Compared with
unless(<F>) versus the complete registered disclosure 'that claim fails if F' with identical claim and falsifier
Tested population
24 frozen complete operational claims across 24 domains, each with a claim-attached observable falsifier
Unit tested
one complete operational claim with its falsifier
How results combine
equal item mean per tokenizer, then least-favourable maximum tokenizer mean; retain the single unless stratum
These are the study author’s declarations. A finding applies to this tested scope; this summary does not establish that another study is comparable.
Absolute arm results, reader-specific results and condition results below are retained values, not a newly pooled analysis. Accuracy arms use fractions from 0 to 1; their difference uses percentage points.
Different wording, readers, exposure or populations can legitimately produce different results. A visible reference is not training the model’s weights. Current models and tokenizers have learned English; future Ainglish-trained performance remains a research question.
