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
-0.59375 tokens per declared item Reported interval: -1.0625 to -0.59375.
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 2 declared conditions
no-undo, can-undo
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
Disagrees with the named original
This eligible row adds one disagreement. An adverse or null direction is a valid result and remains visible.
Is this result within the cost allowance?
This headline is within the allowance.
The reported difference is -0.59375 tokens; the current declaration allows at most 2 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 -0.59375 tokens; the named original reported -0.6875.
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
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 32 English inputs reuse text from either side of the original.
0 of 32 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
no-undo
-1
Not recorded
can-undo
-0.1875
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 32 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
English input
Shredded the receipt; I cannot reverse this.
Ainglish input
Shredded the receipt, no-undo.
Condition
no-undo
Input 2
English input
Sent the notification; I cannot reverse this.
Ainglish input
Sent the notification, no-undo.
Condition
no-undo
Input 3
English input
Broadcast the launch announcement; I cannot reverse this.
Ainglish input
Broadcast the launch announcement, no-undo.
Condition
no-undo
Input 4
English input
Destroyed the expired certificate; I cannot reverse this.
Ainglish input
Destroyed the expired certificate, no-undo.
Condition
no-undo
Input 5
English input
Deleted the sole calibration recording; I cannot reverse this.
Ainglish input
Deleted the sole calibration recording, no-undo.
Condition
no-undo
Input 6
English input
Overwrote the only copy of payroll; I cannot reverse this.
Ainglish input
Overwrote the only copy of payroll, no-undo.
Condition
no-undo
Recorded input digest: 76c29f0de5eecf917d37cc444c20485eb4288cdf4237de9a2983e3c9bfdee4cc
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
marked form (`ACTION, no-undo.` / `ACTION, can-undo(PATH[; HOLDER][; WINDOW][; COST]).`) minus the one fixed careful-English rendering R* v3 (`ACTION; I cannot reverse this.` / `ACTION; I can reverse this via PATH[ within N units][; cost COST].` / `ACTION; HOLDER can reverse this via PATH[...].`), both arms carrying the same ACTION, PATH, HOLDER, WINDOW and COST; renderer noundo_rstar.py sha256 b1cd2787af86de587058fb7914e959a66ddaaedbf974f1f6b440f43832dbeed8
Tested population
the authored 32-pair bank of the row's proposer (bank.json canonical-JSON sha256 f7e05fd81e90786610de559ad3c8ae4d29477180b8051cff20552d1610ef04de, panel-artifacts commit ecab3926b535b8b8ab6326b83b6ed13f24f4687e): 16 no-undo and 16 can-undo, 8 report and 8 instruction per stratum, ACTION word lengths 3:6 4:8 5:8 6:6 7:4, the sixteen can-undo slot combinations on the pinned joint schedule, materialised in profile.json (sha256 bd684a47ec245f1ff265ae35913bf69b06de75d2a6c28d699cbc2125cc02f79b); every ACTION fresh against the 96 prior-bank digests; English arms byte-equal to R* by the packet validator
Unit tested
complete message
How results combine
equal item mean per tokenizer, then maximum tokenizer mean (least-favourable); strata no-undo and can-undo reported at weight 1 each
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.
