LatentEval

INSTRUMENT | eval

Reproducibility Checklist and Manifest Validator

7 cited sources

Check whether your write-up records what someone else would need to rerun your eval: a checklist against the published card schema, and a validator for a pasted or uploaded manifest.

Before you publish a result, check what a second team needs to rerun it. Reproducibility starts with the record: 96.5 percent of evaluation results miss at least one minimal field, checked here against the fields the evaluation card generator publishes.

A complete manifest means the record is complete, never that the number reproduces: a temperature pinned at zero still drifted on two of seven items. Log runs in the run register; see our house methodology.

Mode
Reproducibility checklist
Model name (required)

Check this value.

Check this value.

A family name is not a system

Model version (required)

Check this value.

Check this value.

The resolved snapshot, and the date it resolved (2606.26185)

Temperature (required)

Check this value.

Check this value.

Absent from 93.9 percent of triples (2606.09809)

Max tokens (required)

Check this value.

Check this value.

Absent from 95.6 percent of triples (2606.09809)

Seed (required)

Check this value.

Check this value.

Pinned zero still drifted on 2 of 7 items (2606.26185)

Harness (required)

Check this value.

Check this value.

Check this value.

Named in the minimal sub-schema (2606.09809)

Eval plan (required)

Check this value.

Check this value.

Missing from 100 percent of agentic triples (2606.09809)

Eval limits (required)

Check this value.

Check this value.

Step, tool call, wall clock, or token budget. Missing from 100 percent (2606.09809)

Dataset source (required)

Check this value.

Check this value.

Which set was scored (Croissant)

Dataset version (required)

Check this value.

Check this value.

Sets move, and M-T01 tracks that (Croissant)

Dataset SHA-256 (recommended)

Check this value.

Check this value.

Proves which bytes were scored (Croissant, Data Package)

Dataset size (required)

Check this value.

Check this value.

Items scored, so a reader can size the uncertainty (REFORMS)

Metric name (required)

Check this value.

Check this value.

The metric path is the third leg of the triple (2606.09809)

Metric definition (required)

Check this value.

Check this value.

Two teams score the same output differently

Scoring parser (required)

Check this value.

Check this value.

How a raw output became a score

Judge (required)

Check this value.

Check this value.

Check this value.

Check this value.

Grader config is part of the result (P-19, 2606.26185)

Run count (required)

Check this value.

Check this value.

One run is a sample of one (2606.26185)

Run spread (required)

Read from the card's Result figures, which this checklist does not collect. Record a run-spread figure in the evaluation card generator, export it, then load it here in "Check a manifest". Its live status above still reads from this same rule.

Report variance, never a lone point estimate. A single-run confidence interval is not a spread (2606.26185)

Run date (required)

Check this value.

Check this value.

A provider can move under a fixed version string, so a result has to be datable

Limitations (required)

Check this value.

Check this value.

The weakest section everywhere: 2.1 percent of card text (2401.13822)

Eval set file, for the fingerprint

Up to 10 MB. Nothing is uploaded; the digest is computed in this browser.

Items needing attention

Fill the checklist, or check a manifest, to see a verdict.

Applicable
0
Recorded
0
Invalid
0
Required missing
0
Exempt with a reason
0
Recommended gap
0

Ordered invalid, missing, exempt, recommended gap, recorded.

ItemStatusReason / why it mattersSource
Export

JSON carries the card and re-imports on this page. Markdown is the findings table, for reading in a plain editor. It is one way and is never re-read.

to fill by hand outside this page.

pass one: the published v1 schema (type, range, enum, date, group) · pass two: the reproducibility-minimal overlayHow?

How this is calculated

Two passes, never one. The published evaluation-card v1 schema judges every type, range, enum, date, and grouped field (a judge or a harness needs both a name and a version, or neither) and runs first. A value it rejects reads invalid, never missing and never recorded, and the verdict cannot read complete while one stands. This profile adds no type rule of its own and relaxes none the schema sets.

Only then, requiredness. Of the twenty items, nineteen are required; one, the dataset's SHA-256, is recommended. A value that cleared pass one but was never written down reads missing. An item the reader declares not applicable, with a reason, reads exempt; a blank reason still counts as exempt but prints as a warning, never an error.

Zero is a value. method.temperature: 0 reads recorded, never missing: the most common real-world case a naive presence check gets wrong. A whitespace-only value, or a stand-in like "N/A", "none", or "-", reads missing with a hint pointing at the not-applicable control instead.

Run spread leaves the applicable set only when a run count of exactly one says so. Every other reading of the run count (absent, invalid, two or more) keeps the item applicable, so nothing drops out silently just because a fact this checklist cannot verify was never confirmed.

The fingerprint. SHA-256 over the raw bytes of a picked file, computed with crypto.subtle.digest in this browser: the same value shasum -a 256 prints at a terminal. Compared against a manifest's data.sha256, a mismatch prints a warning naming it, never an error: the two may be different files on purpose.

Formula: pass one: the published v1 schema (type, range, enum, date, group) · pass two: the reproducibility-minimal overlay

The reproducibility-minimal profile

Binds to the published evaluation-card v1 schema, pinned 2026-08-27, never to any TypeScript type. A requiredness change publishes a v2 profile beside this one, never over it.

The twenty reproducibility-minimal v1 items.

ItemP-06 fieldRequiredApplies whenWhy it matters
Model name subject.name required always A family name is not a system
Model version subject.version required always The resolved snapshot, and the date it resolved
Temperature method.temperature required always Absent from 93.9 percent of triples
Max tokens method.maxTokens required always Absent from 95.6 percent of triples
Seed method.seed required the vendor exposes one Pinned zero still drifted on 2 of 7 items
Harness method.harness required a harness or scaffold ran Named in the minimal sub-schema
Eval plan method.evalPlan required agentic run Missing from 100 percent of agentic triples
Eval limits method.evalLimits required agentic run Step, tool call, wall clock, or token budget. Missing from 100 percent
Dataset source data.source required always Which set was scored
Dataset version data.version required always Sets move, and M-T01 tracks that
Dataset SHA-256 data.sha256 recommended a file exists Proves which bytes were scored
Dataset size data.size required always Items scored, so a reader can size the uncertainty
Metric name method.metric required always The metric path is the third leg of the triple
Metric definition task.construct required always Two teams score the same output differently
Scoring parser method.parser required always How a raw output became a score
Judge method.judge required a model graded Grader config is part of the result
Run count method.runs required always One run is a sample of one
Run spread a result.figures row whose intervalKind is 'run-spread' required method.runs recorded and above 1 Report variance, never a lone point estimate. A single-run confidence interval is not a spread
Run date provenance.runOn required always A provider can move under a fixed version string, so a result has to be datable
Limitations limits required always The weakest section everywhere: 2.1 percent of card text

Sources

  1. Evaluation Cards: measuring reproducibility, completeness, provenance and comparability across 101,843 resultsarXiv preprint Retrieved
  2. Temperature control and reproducibility in judge evaluationsarXiv preprint Retrieved
  3. A 2024 study of completeness across 7,433 Hugging Face dataset cardsarXiv preprint (ICLR 2024) Retrieved
  4. The NeurIPS paper checklist's not-applicable-with-a-reason precedentNeurIPS Retrieved
  5. REFORMS, a reporting standard built from one table rather than two documentsREFORMS (Princeton) Retrieved
  6. The MLCommons Croissant metadata spec's per-file SHA-256 conventionMLCommons Retrieved
  7. The Data Package standard's permissive-schema-plus-profile designData Package (Frictionless Data) Retrieved