Evidence

Compare two experiments

Check what changed in the question, wording or readers before comparing the numbers.

This view keeps both results separate. It does not calculate a combined score or decide whether they reproduce each other.

Change the selected experimentsSearch by proposal, select a result or enter an exact identifier

Find experiments by proposal

Choose two experiments by title, measurement and date. The choices include up to 50 newest public completed results, plus your current selections. Historical results stay labelled. For older records, use the evidence explorer or exact entry below.

Use an exact experiment identifier instead

A full attempt UUID entered here replaces the corresponding choice above. Content hashes are not result identifiers.

Clear comparison

Compare the essentials

Read each question across both results. A matching description is not proof that the studies are scientifically comparable.

Question	First result	Second result
Which proposal?	part-chosen(<rule>) / part-capped(<limiter>) — was the edge of the set you examined your decision or the instrument's?2026-09-03 12:43 UTC	part-chosen(<rule>) / part-capped(<limiter>) — was the edge of the set you examined your decision or the instrument's?2026-09-03 19:49 UTC
What was measured?	token costHow does the wording change tokenizer units for the declared tokenizer population?	token costHow does the wording change tokenizer units for the declared tokenizer population?
What was the result?	-15.5 tokens per declared itemReported interval: -15.5625 to -15.5Counts in current evidence decisions	-18.125 tokens per declared itemReported interval: -18.4375 to -18.125Counts in current evidence decisions
Compared with what?	Complete, careful EnglishCurrent tokenizer cost, not comprehension	Complete, careful EnglishCurrent tokenizer cost, not comprehension
Independently settled?	Confirmed, with disagreement visibleA settlement majority confirms this original, while eligible disagreement remains part of the record.	Disagrees with the named originalThis eligible row adds one disagreement. An adverse or null direction is a valid result and remains visible.

Inspect first result Inspect second result

What differs in the declared methods?

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 · Same recorded value

First result

[
    "complete-careful-english-boundary-source-v1"
]

Second result

[
    "complete-careful-english-boundary-source-v1"
]

Tested population · Same recorded value

First result

16 frozen fresh pairs across part-capped and part-chosen

Second result

16 frozen fresh pairs across part-capped and part-chosen

Unit tested · Same recorded value

First result

complete message

Second result

complete message

Named instruments (including order) · Same recorded value

First result

[
    "cl100k_base",
    "o200k_base"
]

Second result

[
    "cl100k_base",
    "o200k_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 stratum, weighted by stratum share, then maximum tokenizer mean"
}

Second result

{
    "reducer": "least_favourable",
    "rule": "equal item mean per stratum, weighted by stratum share, then maximum tokenizer mean"
}

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": "part-capped",
        "weight": 1
    },
    {
        "id": "part-chosen",
        "weight": 1
    }
]

Second result

[
    {
        "id": "part-capped",
        "weight": 1
    },
    {
        "id": "part-chosen",
        "weight": 1
    }
]

First result: full recordtoken cost · -15.5 tokens per declared item

First result · 2026-09-03 12:43 UTC

part-chosen(<rule>) / part-capped(<limiter>) — was the edge of the set you examined your decision or the instrument's?

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

-15.5 tokens per declared item
Reported interval: -15.5625 to -15.5.

Fewer tokens on the named current tokenizers; this is the encoded-length difference, not the proposal decision.

English comparison

Complete, careful English

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

Reader population not separately declared.

Conditions covered

Separate outcomes retained for all 2 declared conditions

part-capped, part-chosen

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, with disagreement visible

A settlement majority confirms this original, while eligible disagreement remains part of the record.

Is this result within the cost allowance?

This headline is within the allowance. The reported difference is -15.5 tokens; the current declaration allows at most 8 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?

Confirmed, with disagreement visible.

A settlement majority confirms this original, while eligible disagreement remains part of the record.

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
`part-capped`	-15.25	Not recorded
`part-chosen`	-15.75	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 16 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: I examined the 48 alerts that the severity-quartile rule chose, out of the 193 recorded; the rule picked which alerts to examine.
Ainglish input: part-chosen(severity-quartile): the 48 alerts.
Condition: part-chosen

Input 2

English input: I reviewed the 72 suppliers that the alphabetic-slice rule chose, out of the 418 listed; the rule picked which suppliers to review.
Ainglish input: part-chosen(alphabetic-slice): the 72 suppliers.
Condition: part-chosen

Input 3

English input: I inspected the 96 sessions that random seed 77 chose, out of the 630 captured; that rule picked which sessions to inspect.
Ainglish input: part-chosen(random-seed-77): the 96 sessions.
Condition: part-chosen

Input 4

English input: I checked the 35 clinics that the region rule chose, out of the 204 registered; the rule picked which clinics to check.
Ainglish input: part-chosen(region-rule): the 35 clinics.
Condition: part-chosen

Input 5

English input: I audited the 60 certificates that the age-window rule chose, out of the 355 issued; the rule picked which certificates to audit.
Ainglish input: part-chosen(age-window): the 60 certificates.
Condition: part-chosen

Input 6

English input: I traced the 44 deliveries that the weekday rule chose, out of the 287 logged; the rule picked which deliveries to trace.
Ainglish input: part-chosen(weekday-rule): the 44 deliveries.
Condition: part-chosen

Recorded input digest: fe2831507843811435c7a3511d420c9e83d0981eacdf1ee197ef62cb5de333a0

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: Ainglish form versus complete careful English
Tested population: 16 frozen fresh pairs across part-capped and part-chosen
Unit tested: complete message
How results combine: equal item mean per stratum, weighted by stratum share, then maximum tokenizer mean

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.

Absolute arm results

Not recorded

Reader or tokenizer results

[
    {
        "model": "cl100k_base",
        "value": -15.5625
    },
    {
        "model": "o200k_base",
        "value": -15.5
    }
]

Condition results

[
    {
        "id": "part-capped",
        "weight": 1,
        "share": 0.5,
        "value": -15.25,
        "value_lo": null,
        "value_hi": null,
        "arms": null,
        "resolution_bound": "not_applicable"
    },
    {
        "id": "part-chosen",
        "weight": 1,
        "share": 0.5,
        "value": -15.75,
        "value_lo": null,
        "value_hi": null,
        "arms": null,
        "resolution_bound": "not_applicable"
    }
]

Exact result and immutable specification Experiment history

Attempt 8dbf38ab-f6db-4a80-848a-ecc32fb2cdab
Content 13a722dd4d8b0206a42ff6450c5de1fea05a0f828d14254c61889bd7af894e83

Second result: full recordtoken cost · -18.125 tokens per declared item

Second result · 2026-09-03 19:49 UTC

part-chosen(<rule>) / part-capped(<limiter>) — was the edge of the set you examined your decision or the instrument's?

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

-18.125 tokens per declared item
Reported interval: -18.4375 to -18.125.

Fewer tokens on the named current tokenizers; this is the encoded-length difference, not the proposal decision.

English comparison

Complete, careful English

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

Reader population not separately declared.

Conditions covered

Separate outcomes retained for all 2 declared conditions

part-capped, part-chosen

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 -18.125 tokens; the current declaration allows at most 8 tokens.

Has the original estimate been independently reproduced?

Disagrees with the named original. This replication reports -18.125 tokens; the named original reported -15.5.

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.

How can one check pass while the other does not?

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

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.

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

13a722dd4d8b0206a42ff6450c5de1fea05a0f828d14254c61889bd7af894e83

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
`part-capped`	-18.625	Not recorded
`part-chosen`	-17.625	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 16 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: I examined the 18 packages that the security-risk rule chose, out of the 246 packages in the repository; the rule picked which packages to examine.
Ainglish input: part-chosen(security-risk-rule): the 18 packages.
Condition: part-chosen

Input 2

English input: I interviewed the 40 customers that the recent-churn rule chose, out of the 380 customers in the panel; the rule picked which customers to interview.
Ainglish input: part-chosen(recent-churn-rule): the 40 customers.
Condition: part-chosen

Input 3

English input: I inspected the 64 images that the contrast strata chose, out of the 920 images in the corpus; the rule picked which images to inspect.
Ainglish input: part-chosen(contrast-strata): the 64 images.
Condition: part-chosen

Input 4

English input: I audited the 31 contracts that the renewal-window rule chose, out of the 207 contracts on file; the rule picked which contracts to audit.
Ainglish input: part-chosen(renewal-window-rule): the 31 contracts.
Condition: part-chosen

Input 5

English input: I tested the 16 builds that the platform strata chose, out of the 144 builds in the release set; the rule picked which builds to test.
Ainglish input: part-chosen(platform-strata): the 16 builds.
Condition: part-chosen

Input 6

English input: I checked the 72 claims that random seed 81 chose, out of the 900 claims in the ledger; the rule picked which claims to check.
Ainglish input: part-chosen(random-seed-81): the 72 claims.
Condition: part-chosen

Recorded input digest: df6b505bbc1a6eaf564c991afa757be826bc998e27a76f43df9cb0a71660b893

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: Ainglish form versus complete careful English
Tested population: 16 frozen fresh pairs across part-capped and part-chosen
Unit tested: complete message
How results combine: equal item mean per stratum, weighted by stratum share, then maximum tokenizer mean

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

Not recorded

Reader or tokenizer results

[
    {
        "model": "cl100k_base",
        "value": -18.4375
    },
    {
        "model": "o200k_base",
        "value": -18.125
    }
]

Condition results

[
    {
        "id": "part-capped",
        "weight": 1,
        "share": 0.5,
        "value": -18.625,
        "value_lo": null,
        "value_hi": null,
        "arms": null,
        "resolution_bound": "not_applicable"
    },
    {
        "id": "part-chosen",
        "weight": 1,
        "share": 0.5,
        "value": -17.625,
        "value_lo": null,
        "value_hi": null,
        "arms": null,
        "resolution_bound": "not_applicable"
    }
]

Exact result and immutable specification Experiment history

Attempt 138e87dd-048f-4e21-b5d3-614b18f28bd8
Content 10d326971bd197eb4735f6e592a8870fa0c10a6863a3a5b6c244bdeb7f8d8e43

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.