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
-7.125 tokens per declared item Reported interval: -10 to -7.125.
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
falsum-ref
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
falsum-ref
-7.125
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 · reef-oxygen
English input
The claim reef-oxygen-stable is refuted by the observation named buoy-DO-17, whose observable delta is: dissolved oxygen fell from 7.1 to 4.8 mg/L; re-read buoy DO-17 channel 2.
Ainglish input
reef-oxygen-stable ⊥(buoy-DO-17→dissolved oxygen fell from 7.1 to 4.8 mg/L; re-read buoy DO-17 channel 2)
Condition
falsum-ref
Input 2 · rail-clearance
English input
The claim tunnel-clearance-valid is refuted by the observation named laser-scan-L44, whose observable delta is: the minimum clearance changed from 310 to 274 mm; rerun profile L44 at chainage 8.2 km.
Ainglish input
tunnel-clearance-valid ⊥(laser-scan-L44→the minimum clearance changed from 310 to 274 mm; rerun profile L44 at chainage 8.2 km)
Condition
falsum-ref
Input 3 · archive-fixity
English input
The claim archive-bundle-intact is refuted by the observation named sha256-audit, whose observable delta is: object 6F now hashes to 91c2 rather than 77ad; hash object 6F from the sealed copy.
Ainglish input
archive-bundle-intact ⊥(sha256-audit→object 6F now hashes to 91c2 rather than 77ad; hash object 6F from the sealed copy)
Condition
falsum-ref
Input 4 · trial-balance
English input
The claim trial-balance-equal is refuted by the observation named ledger-query-Q9, whose observable delta is: debits exceed credits by 420 GBP; rerun saved query Q9 on journal 2026-09.
Ainglish input
trial-balance-equal ⊥(ledger-query-Q9→debits exceed credits by 420 GBP; rerun saved query Q9 on journal 2026-09)
Condition
falsum-ref
Input 5 · cold-room
English input
The claim cold-room-compliant is refuted by the observation named thermistor-T12, whose observable delta is: temperature rose from 3.4 to 9.2 C for 18 minutes; replay logger T12 interval 02:10-02:28.
Ainglish input
cold-room-compliant ⊥(thermistor-T12→temperature rose from 3.4 to 9.2 C for 18 minutes; replay logger T12 interval 02:10-02:28)
Condition
falsum-ref
Input 6 · orbital-clock
English input
The claim orbital-clock-synced is refuted by the observation named telemetry-frame-812, whose observable delta is: clock offset increased from 4 to 83 ms; decode field 19 in frame 812.
Ainglish input
orbital-clock-synced ⊥(telemetry-frame-812→clock offset increased from 4 to 83 ms; decode field 19 in frame 812)
Condition
falsum-ref
Recorded input digest: f9be4cf873fb43f02462403e95ce003f270a95f1520460282dc8794783030df1
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
<claim> ⊥(<instrument>→<delta>) versus the registered complete English mapping with identical claim, instrument and load-bearing observable delta
Tested population
24 frozen complete refutation notices across 24 domains; each names a claim, instrument, observed change and executable re-check path
Unit tested
one complete claim/instrument/delta refutation notice
How results combine
equal item mean per tokenizer, then least-favourable maximum tokenizer mean; retain the single falsum-ref 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.
