TETRACTA · MODEL X-RAY

Before / after comparison

Two artifacts compared under one fixed measurement contract
Depth view · Decision, output status, available verification evidence and categorical depth observations
VG1 · betars-1.7 · rp-1.3
A · SOURCE
Qwen/Qwen2.5-7B
@ d149729398750b98c0af14eb82c78cfe92750796
Reference artifact
→
B · AFTER
Qwen/Qwen2.5-7B
@ d149729398750b98c0af14eb82c78cfe92750796
Candidate artifact
○
VALIDATION-PENDING DESCRIPTIVE RESULT · no calibrated risk or pass/fail grade
No difference observed
No evaluated response difference met the versioned instrument decision rule under this exact comparison contract.
Internal response
Not observed
Response difference on the evaluated set
Text output
Unchanged on sampled inputs
All evaluated generated continuations match under exact token-sequence comparison. The comparison uses the same tokenizer and greedy-generation contract; it is not a semantic-quality judgment.
Repeat check
2 / 2
Matching versioned measurement payloads reported from fresh processes within one allocation.

Included summaries Content and scope, not a quality or risk grade

Record or summaryThis report
Measurement summaryIncluded
Repeat-check summaryIncluded · 2 / 2 matching measurement payloads reported
Same-artifact control summaryIncluded · the same 2 measurements
Underlying recordsRetained in the VG1 archive; not included in this copy
Operational receiptNot included · unsigned review copy
Statistical noise envelopeNot certified

Repeat and control details

Repeat checkThis completed allocation was configured for 2 repeat measurements; all 2 completed and their versioned measurement payloads matched byte for byte. The payload contains all versioned measurement fields. Execution identifiers, runtime records and timing metadata are recorded separately; they are not part of this equality claim. Fresh processes and model loads within one allocation; this does not establish agreement across devices or allocations. Counts refer to this allocation, not the full development history. No preregistration claim is made.
Same-artifact controlThe 2 measurements above are themselves same-artifact comparisons. They supply both the repeat and control observations; they are counted once, not as additional executions. 2 / 2 recorded no internal finding and matching token sequences. These are observed controls, not a calibrated noise envelope or an error-rate estimate.
Operational receiptNo customer-verifiable signed receipt is included in this HTML copy. No signature or deletion claim is made.
Noise-floor statusNo model-specific statistical noise envelope is certified by this report.

What this means for you

MeasuredThe instrument found no response difference on this fixed input set. This does not establish that the artifacts are equivalent on other inputs or tasks.
Text outputThe output-text comparison did not observe a change.
PriorityRetain this as a bounded reference. Neither comparison found a change on its evaluated set. Use the pinned record for later comparisons; it does not establish equivalence on unseen inputs.

How the finding was decided

The reported internal finding uses the fixed, thresholded decision rule recorded for mv-1.4; it is not a test for any non-zero floating-point value. This rule is not, by itself, a measured noise floor for this model.

Text output is evaluated separately under exact greedy token-sequence comparison. The outcome is on the fixed input set, not a population rate or a capability score.

Structural coverage: 24 contributing probes per evaluated position. This is neither the number of output continuations nor a count of independent tasks.

Observed responses through the model

No difference was observed at any evaluated position under this contract.

embedding output: not evaluatedEmbdecoder block 1 output: not evaluateddecoder block 2 output: not evaluateddecoder block 3 output: not evaluateddecoder block 4 output: no difference observed4decoder block 5 output: no difference observeddecoder block 6 output: no difference observeddecoder block 7 output: no difference observeddecoder block 8 output: no difference observed8decoder block 9 output: no difference observeddecoder block 10 output: no difference observeddecoder block 11 output: no difference observeddecoder block 12 output: no difference observed12decoder block 13 output: no difference observeddecoder block 14 output: no difference observeddecoder block 15 output: no difference observeddecoder block 16 output: no difference observed16decoder block 17 output: no difference observeddecoder block 18 output: no difference observeddecoder block 19 output: no difference observeddecoder block 20 output: no difference observed20decoder block 21 output: no difference observeddecoder block 22 output: no difference observeddecoder block 23 output: no difference observeddecoder block 24 output: no difference observed24decoder block 25 output: no difference observeddecoder block 26 output: no difference observeddecoder block 27 output: no difference observedfinal normalized output: no difference observedFinal
difference observedno difference observednot evaluated

0 of 25 evaluated positions show a difference. This count describes the reach of the observed response, not its size or the number of changed weights.

Recorded outputObservation
embedding output through decoder block 3 outputnot evaluated
decoder block 4 output through decoder block 27 outputno difference observed
final normalized outputno difference observed

Blocks are numbered from 1. Final is the output after the last normalization; the last block before normalization is not separately displayed. Not evaluated means no observation was made there; it is not a below-threshold result.

A difference can propagate from one output to later outputs. This view does not locate edited weights, identify a cause or show which capabilities changed. Controlled small-model findings are not a validation result for this model or execution environment.

What differs between A and B

Artifact metadataRecorded comparison
Model configuration semanticsmatched
Tokenizer semanticsmatched
Generation settings semanticsmatched

Recorded relation: combined artifact change. Do not attribute the finding to fine-tuning or weights alone.

Matched refers only to the recorded semantic summaries. It does not mean the complete artifacts are identical. Different configurations, tokenizers or generation settings can matter alongside weight changes.

Recommended next checks Guidance, not a measurement

  1. Control the comparison contractReview the metadata differences above. Establish comparable tokenizer and generation settings before interpreting an output comparison; preserve each artifact revision.
  2. Run the relevant regression setCompare A and B on the tasks your users rely on, including output format and failure cases. Set acceptance criteria before inspecting the results.
  3. Keep an auditable decisionRetain this report with the task results and deployment revision. Separate the measured observation from your release decision.

Provenance and record

InstrumentVG1
Measurement contractmv-1.4
Report schema / presentationrs-1.7 / rp-1.3

No operational receipt is included in this copy. No signing or deletion claim is made.

Recorded context

Pin these fields when you diff this report against a future scan. Matching context is a prerequisite for interpretation, not proof of the cause of a difference. These recorded public identities help identify the input artifacts. This review copy is unsigned; no customer verification receipt is attached.

Run window (UTC)2026-09-08 11:27:22 UTC to 2026-09-08 11:38:09 UTC
Model referenceQwen/Qwen2.5-7B → Qwen/Qwen2.5-7B
Artifact evidenceArtifact-identity and measurement records are preserved in the VG1 archive; this HTML contains no model weights
HF revisionA: d149729398750b98c0af14eb82c78cfe92750796 · B: d149729398750b98c0af14eb82c78cfe92750796
Probe set / scan configps-1.1 / sc-1.0
Reference setno release reference set; estimator validation pending
Report schemars-1.7
Compute isolationmanaged compute; infrastructure identity is withheld
Runtime envelopefixed release configuration; bitwise equality across device types is not guaranteed

Reading this report alongside another scan

Check the artifact pair, revisions and measurement contract first. The categorical findings describe each recorded comparison; they do not rank different models. A different contract or execution environment needs its own comparability evidence. No numerical severity score is supplied.

Use as a checkpoint reference

For supported checkpoint comparisons, retain the starting revision, this measurement contract and the output settings with your training record. Compare the next checkpoint against that same reference, then retain its task results alongside the scan.

For the next checkpointRecord alongside this scan
Reference and candidateExact repository revisions
Comparable measurementSame instrument, input contract and approved execution profile
Training changeYour fine-tune, merge or pruning record
Adoption decisionYour task acceptance results and deployment revision

A reference record supports change tracking. It does not define an optimal model or a target score for training. Simulation results describe the simulated state; actual checkpoint changes require their own before/after comparison.

Your findings and the protected instrument

Delivered to you

The observation, output-text status, artifact context and, where applicable, metadata comparison, plus provenance and categorical depth detail when included.

Kept private

Raw profiles, per-item records, probe construction, thresholds, transformations and internal traces. The report contains no model weights, credentials or infrastructure identifiers.

Honesty box — scope and limits

This descriptive result does not identify cause, capability, quality, safety or deployment impact. It is not a task-quality measurement. Do not target a layer or make a shipping decision from this report alone. No calibrated risk or pass/fail grade is assigned. No cross-model ranking or cross-device bitwise-equality claim is made.

Historical results under superseded estimators do not validate this release. The categorical depth view, when included, describes recorded responses; it is not a diagnosis of edited weights.