feat(perception): stabilize pre-capture methodology
This commit is contained in:
DCCONSTRUCTIONS
@@ -3,7 +3,25 @@
|
||||
This plan supersedes the app-dependent experiment order in the reference Bible.
|
||||
Each gate produces evidence and an explicit GO, PAUSE or BLOCKED result.
|
||||
|
||||
## Current checkpoint — 2026-07-24
|
||||
## Architecture stabilization checkpoint — 2026-07-28
|
||||
|
||||
| Gate | Current result |
|
||||
| --- | --- |
|
||||
| RAVNOVES00 methodology | CORRECTED — E41 proves E37 labels are engineering evidence, not independent truth, and reproduces connected development/evaluation overlap. The 146-row slice is historical evaluated visible validation. |
|
||||
| Prediction/evaluation boundary | GO — truth-free E41 predictor package/result are physically separate from E37 and reproduce E40 predictions exactly; the visible evaluator runs only after inference. |
|
||||
| Source-scoped quality | PAUSE — E40 visible engineering evaluation is 84.2466% presence, 84.2466% geometry association and 94.5205% freshness, with 11 high-severity failures, complete accounting and zero false-free claims. This is not a blind accuracy gate. |
|
||||
| Structural regressions | GO (bounded) — E42 passes predictor identity/order/chunk invariance and PointSlab order/SE(3) checks. Raw-producer and cross-route invariance remain unproved. |
|
||||
| Native pipeline telemetry | GO for producer/normalizer contract — lifecycle events and JSONL/MQTT sink boundaries exist and an E41 smoke run records real stage accounting. Durable Worker 006 MQTT wiring/deployment remains pending. |
|
||||
| Evidence storage | MEASURED — E44 finds 525,471,092 logical bytes, 312,753,179 unique-content bytes and 1.680146× amplification across 14 E30–E40 roots. Exact-content references/deduplication precede any format migration. |
|
||||
| Future transfer | PREREGISTERED — E43 freezes same-K1/mount/calibration/firmware, required streams, connected-component split and independent label reveal. Capture and labels do not yet exist. |
|
||||
| Product interface | DEFERRED — no new windows, page anatomy or design changes are part of this stabilization increment. |
|
||||
|
||||
The governing decision is
|
||||
[`ADR 0032`](adr/0032-ravnoves00-methodology-correction-before-transfer.md).
|
||||
Historical E37–E40 artifacts remain immutable; only the claims made from them
|
||||
change.
|
||||
|
||||
## Earlier checkpoint — 2026-07-24
|
||||
|
||||
| Stage | Result |
|
||||
| --- | --- |
|
||||
|
||||
@@ -8,7 +8,8 @@ explainability implemented; L2.6e recorded-source-paced bounded shadow
|
||||
qualified; E28 complete worker replay accepted; E29 camera-first semantic and
|
||||
parallel geometry-only replay implemented; E30–E35 source-scoped qualification
|
||||
accepted; RAVNOVES00 reference-source product maturation active; E36 transfer
|
||||
deferred by ADR 0030
|
||||
preregistered and deferred by ADR 0030/0032; E41 methodology boundary, E42
|
||||
metamorphic checks and E44 amplification audit complete
|
||||
Scope: passively received real-time K1 point/pose evidence, immutable replay and
|
||||
future live shadow processing
|
||||
Explicitly out of scope: K1 firmware modification, a new onboard exporter, new
|
||||
@@ -652,6 +653,9 @@ to its source session under that audit profile. The transfer replay was not
|
||||
created. ADR 0030 subsequently deferred TEST007 and E36 as current priorities:
|
||||
RAVNOVES00 source-scoped quality, product workflow and a reference release
|
||||
candidate now come first. No new capture is required or authorized by default.
|
||||
ADR 0032 retains that priority while preregistering the later same-K1/new-route
|
||||
capture under E43. The protocol is frozen; no capture or labels currently
|
||||
exist.
|
||||
The active gates are defined in
|
||||
`docs/20_RAVNOVES00_REFERENCE_SOURCE_PRODUCT_PLAN.md`.
|
||||
|
||||
@@ -719,17 +723,37 @@ allowed to change; E31 determined the E32 profile; E32 determined the E33
|
||||
runtime input; E32/E33 then bound the E34 layer; E32–E34 then bound E35.
|
||||
Exact nominal and degradation accounting are closed, so A8 is complete.
|
||||
|
||||
R1 perception quality is now measured by two immutable Worker 006 results.
|
||||
R1 perception quality was initially measured by two immutable Worker 006 results.
|
||||
E38 established the baseline at 82.2% presence, 81.5% geometry association,
|
||||
95.2% freshness and 14 high-severity failures. E39 result
|
||||
`e39-perception-refinement-2fd253940c9d9a2fd3b3237f3f0932f81a9f69741935d793771243d5779af464`
|
||||
used a fixed package-bound camera + LiDAR feature projection and improved
|
||||
presence and geometry association to 84.9% while reducing high-severity
|
||||
failures to 8. Freshness remained above target at 93.2%. Its development
|
||||
cross-validation exceeded 90% but did not predict the sealed result, so R1
|
||||
remains open. The next iteration is development-only grouped time/scene
|
||||
qualification with a source-coordinate-free representation; validation is not
|
||||
opened for item-level tuning.
|
||||
cross-validation exceeded 90% but did not predict the evaluated slice, so R1
|
||||
remains open.
|
||||
|
||||
The E41 source audit corrects the meaning of those measurements. E37's
|
||||
`independent_ground_truth=false` provenance, connected group overlap and
|
||||
prediction/evaluation co-location mean the 146 cases were historically
|
||||
evaluated visible engineering labels, not a sealed blind set. E40 package-bound
|
||||
result
|
||||
`e40-perception-product-gate-e96eec9fd68c3ffaaee898d46285dd329191267200011680f084c75095b92e9a`
|
||||
reached 84.2466% presence, 84.2466% geometry association and 94.5205% freshness
|
||||
with 11 high-severity failures, complete accounting and zero false-free claims.
|
||||
It remains useful source-scoped engineering evidence and does not close R1.
|
||||
|
||||
E41 result
|
||||
`e41-methodology-audit-703dac176f20843ad6fbb9a1060b199942fba05e3bb34d3e2264ee87c7bfdd96`
|
||||
reproduces `14` exact-frame, `38` track, `58` fixed-time-block and `59`
|
||||
connected whole-track-or-scene overlaps. The replacement predictor package and
|
||||
prediction contain no reference/split/scoring material, reproduce E40
|
||||
predictions exactly and are evaluated only afterward by
|
||||
`e41-visible-evaluation-57aa6c8569e8339630406e3c88cd539df13917fb08c1af57f0baae5cdd1069d2`.
|
||||
E42 adds bounded structural invariants; E44 measures exact-content
|
||||
amplification; E43 freezes the later independent transfer protocol. ADR 0032
|
||||
contains the corrected decision boundary and current priority order.
|
||||
|
||||
E36 is the first generalization gate. A separate product decision follows:
|
||||
either keep the result as operator/shadow evidence, or start L5 occupied-space
|
||||
integration. No LAB in this cycle can enable navigation, commands or safety
|
||||
|
||||
@@ -32,12 +32,24 @@ candidate is accepted:
|
||||
|
||||
- new perception and product iterations use RAVNOVES00;
|
||||
- each iteration creates a new immutable LAB and never rewrites an older one;
|
||||
- TEST007, E36 and new physical collection are deferred, not required gates;
|
||||
- TEST007 and E36 replay are deferred, not required current gates;
|
||||
- no agent may initiate or imply a new capture without explicit operator
|
||||
authorization;
|
||||
- the active quality target is at least `90%` on each frozen task-level
|
||||
validation dimension, never one unqualified aggregate accuracy number.
|
||||
|
||||
ADR 0032 corrects the present validation meaning. E37's 146-row slice is
|
||||
historical evaluated visible engineering validation: its labels are not
|
||||
independent truth and its development/evaluation groups overlap. It remains an
|
||||
immutable comparison contract, but it cannot close a blind accuracy gate.
|
||||
Prediction packages must exclude truth, split, severity and scoring material,
|
||||
and evaluation must join those artifacts only after prediction.
|
||||
|
||||
The owner has authorized preparation for a later same-K1/new-route recording,
|
||||
not agent-initiated collection. E43 freezes that future protocol before data
|
||||
exists. Until the recording is supplied, no LAB result, metric or transfer
|
||||
claim is created for it.
|
||||
|
||||
This policy is a development priority, not a universality claim. Later
|
||||
second-source transfer remains required before any cross-route or cross-camera
|
||||
generalization statement.
|
||||
|
||||
@@ -1,8 +1,8 @@
|
||||
# RAVNOVES00 reference-source product plan
|
||||
|
||||
Date: 2026-07-27
|
||||
Date: 2026-07-28
|
||||
|
||||
Status: active
|
||||
Status: active; methodology corrected by ADR 0032
|
||||
|
||||
ADR 0030 makes RAVNOVES00 the sole physical reference source for the current
|
||||
Mission Core product-maturation cycle. This plan turns that decision into
|
||||
@@ -24,13 +24,20 @@ missions. Mission Core should answer:
|
||||
The product is not being designed for adversarial, tactical or continuously
|
||||
novel environments.
|
||||
|
||||
## Acceptance meaning
|
||||
## Acceptance meaning after source audit
|
||||
|
||||
`90%` is a target for reviewed task-level correctness on RAVNOVES00, not a
|
||||
generic marketing accuracy score.
|
||||
|
||||
The next acceptance contract must freeze a validation set before further
|
||||
tuning and measure at least:
|
||||
E37 froze a useful source-scoped engineering contract, but E41 proved that its
|
||||
labels are not independent ground truth and its development/validation rows
|
||||
share frames, tracks and time/scene components. The existing 146-row slice is
|
||||
therefore **historical evaluated visible validation**, not a sealed blind set.
|
||||
Its metrics remain useful for immutable method comparison, but cannot prove
|
||||
independent perception accuracy.
|
||||
|
||||
The next independent acceptance contract must freeze a leakage-free validation
|
||||
set before labels are revealed and measure at least:
|
||||
|
||||
| Dimension | Question | Source-scoped target |
|
||||
| --- | --- | ---: |
|
||||
@@ -44,30 +51,35 @@ Targets apply separately. A strong presence score cannot compensate for bad
|
||||
geometry association or hidden stale evidence.
|
||||
|
||||
The reviewed denominator, sampling strata and severity classes must be frozen
|
||||
before the next tuning cycle. Development items may be used for diagnosis and
|
||||
changes; sealed validation items may only be used for evaluation. High-impact
|
||||
before the next independent tuning cycle. Connected scene/track/time components,
|
||||
not individual item hashes, are the minimum partition unit. Development items
|
||||
may be used for diagnosis and changes; independent validation labels remain
|
||||
unavailable to the predictor until frozen predictions exist. High-impact
|
||||
failures remain blocking even when an aggregate percentage passes.
|
||||
|
||||
## Current path
|
||||
|
||||
| Gate | Deliverable | Exit |
|
||||
| --- | --- | --- |
|
||||
| R0 — acceptance contract | Freeze task ontology, reviewed denominator, development/validation split, metrics, severity and error budget over RAVNOVES00 | Reproducible evaluation with no post-result denominator changes |
|
||||
| R0 — acceptance methodology | Preserve the E37 visible engineering contract, enforce predictor/evaluator separation and preregister an independent, leakage-free future split | No truth material in the predictor; connected-component split; labels revealed only after frozen prediction |
|
||||
| R1 — perception quality | Iterate detector, camera↔geometry association, source-time handling and conservative corrections as new immutable LABs | Each task-level validation dimension reaches `>= 90%` or has an explicit bounded exception |
|
||||
| R2 — temporal product state | Refine current/held/stale/unknown presentation and bounded occupied telemetry without polluting the persistent reconstruction | Stable task-relevant state, closed accounting and deterministic degradation |
|
||||
| R3 — product workflow | Complete reusable source/LAB selection, method summary, visual evidence, comparison, controls and conclusion templates | A non-expert can understand the run; an engineer can inspect exact provenance and metrics |
|
||||
| R3 — product workflow | Separate owner-reviewed interface phase; no new windows, page anatomy or design changes in the present stabilization increment | Product brief reviewed separately before implementation |
|
||||
| R4 — reference release candidate | Replay the complete RAVNOVES00 source through the accepted pipeline at recorded pace with regression, resource and recovery evidence | Source-scoped release decision for the known-location product envelope |
|
||||
| R5 — later transfer | Record a new route/camera only after explicit operator authorization and run the frozen release candidate unchanged | Generalization decision; no retuning before comparison |
|
||||
| R5 — later transfer | Under the preregistered E43 protocol, record the same K1 and mount over a new route after explicit operator authorization; freeze predictions before independent labels are revealed | Generalization decision; no retuning before comparison |
|
||||
|
||||
R0–R4 are the active path. R5 is intentionally deferred.
|
||||
R0–R2 and non-UI R4 preparation are the active stabilization path. R3 is
|
||||
explicitly deferred for separate discussion. R5 has a frozen protocol but no
|
||||
capture or labels yet.
|
||||
|
||||
## Current evidence
|
||||
|
||||
R0 is closed by immutable result
|
||||
E37 created immutable result
|
||||
`e37-ravnoves-acceptance-01b1efd586f747341c712d82f0907b39436a6f91ae92b1dfae987eca05fd8344`.
|
||||
It freezes 486 reviewed RAVNOVES00 items as 340 development and 146 sealed
|
||||
validation cases, with separate presence, geometry-association and freshness
|
||||
references.
|
||||
It freezes 486 reviewed RAVNOVES00 items as 340 development and 146 visible
|
||||
evaluation cases, with separate presence, geometry-association and freshness
|
||||
references. Its own provenance records `484` engineering-reviewed items, `2`
|
||||
human exceptions and `independent_ground_truth=false`.
|
||||
|
||||
The first R1 measurement is immutable result
|
||||
`e38-perception-baseline-a272f82988cd9a7e071fad94c3e9fb49daf804fdcca523f853445fd3113a62b1`.
|
||||
@@ -84,15 +96,59 @@ Its fixed robust three-neighbour refinement used exact camera-crop, LiDAR
|
||||
shape and projection features and was selected only through five-fold
|
||||
development cross-validation. Development reached 90.3% presence, 90.6%
|
||||
geometry association and 96.5% freshness without validation-label access.
|
||||
Sealed validation reached 84.9%, 84.9% and 93.2% respectively. Accounting
|
||||
The then-designated validation slice reached 84.9%, 84.9% and 93.2%
|
||||
respectively. Accounting
|
||||
remains 100%, false-free claims remain zero and high-severity failures fall
|
||||
from 14 to 8, but R1 is still not accepted.
|
||||
|
||||
The next R1 iteration must address the development-to-validation gap before
|
||||
adding model complexity: grouped time/scene development folds replace
|
||||
item-hash folds, and a source-coordinate-free representation is compared under
|
||||
that harder protocol. No individual E39 validation label may be used for
|
||||
diagnosis, fitting or selection.
|
||||
E40 package-bound result
|
||||
`e40-perception-product-gate-e96eec9fd68c3ffaaee898d46285dd329191267200011680f084c75095b92e9a`
|
||||
used `89` camera-only development rows over `125` feature dimensions (`90`
|
||||
variable and `35` constant). Against the already-visible E37 evaluation slice
|
||||
it reached `84.2466%` presence, `84.2466%` geometry association and
|
||||
`94.5205%` freshness, with `11` high-severity failures, complete accounting and
|
||||
zero false-free claims. It does not pass R1.
|
||||
|
||||
E41 then audited the implementation and artifacts together. Immutable audit
|
||||
`e41-methodology-audit-703dac176f20843ad6fbb9a1060b199942fba05e3bb34d3e2264ee87c7bfdd96`
|
||||
found overlap across `14` exact frames, `38` track identities, `58` fixed
|
||||
50-frame time blocks and `59` connected whole-track-or-scene groups. It also
|
||||
proved that E40 prediction and evaluation material were co-located. R0 is
|
||||
therefore reopened for blind acceptance.
|
||||
|
||||
The replacement boundary is executable:
|
||||
|
||||
- predictor package
|
||||
`e41-predictor-package-fb35f42698013c63c6d417fd3b26986e9d3b89d12dce138bab18fe4203d593ad`
|
||||
contains no reference, split, severity, scoring or truth material;
|
||||
- prediction result
|
||||
`e41-predictions-adcc1671de44646145a396d5815e522d6b4257d6ead9ace995761751655b3aac`
|
||||
reproduces the physical E40 predictions exactly;
|
||||
- separate visible evaluation
|
||||
`e41-visible-evaluation-57aa6c8569e8339630406e3c88cd539df13917fb08c1af57f0baae5cdd1069d2`
|
||||
reproduces the E40 engineering metrics and blockers only after inference.
|
||||
|
||||
E42 immutable result
|
||||
`e42-metamorphic-suite-c358204e0b9e10eea5871e90b7c4fbe2c132f1f6d87074b2c5ffec18eaeebb40`
|
||||
passes predictor identity, row-order and chunk-boundary invariance, metadata
|
||||
absence, PointSlab row-order invariance and rigid-coordinate equivariance. It
|
||||
does not prove raw-producer or cross-route invariance.
|
||||
|
||||
E44 immutable audit
|
||||
`e44-data-amplification-89791978894ec785009e5b76c9325de4d9e910237af57476c9331eb47a8ccca4`
|
||||
measures `525,471,092` logical bytes versus `312,753,179` unique-content bytes
|
||||
across 14 admitted E30–E40 roots: `1.680146×` amplification and
|
||||
`212,717,913` exact duplicate bytes. The immediate storage action is
|
||||
content-addressed referencing and exact deduplication, not an unmeasured
|
||||
format/database migration.
|
||||
|
||||
E43 immutable protocol
|
||||
`e43-future-capture-protocol-28f091b9648daffce988d44c183e21f56d77988061630de934f8003fb13701d8`
|
||||
preregisters the later same-K1/new-route transfer. It requires the same mount,
|
||||
calibration and firmware; `480–900 s` duration; at least `60 s` control bridge
|
||||
and `360 s` new route; camera, registered LiDAR, pose and pipeline telemetry;
|
||||
connected component partitioning; two independent reviewers; and label reveal
|
||||
only after frozen prediction. No such capture or labels currently exist.
|
||||
|
||||
## Experiment rules
|
||||
|
||||
@@ -109,15 +165,28 @@ diagnosis, fitting or selection.
|
||||
7. No run claims universal accuracy, planner fitness, navigation or safety
|
||||
acceptance merely because its source-scoped target passes.
|
||||
|
||||
## Deferred work
|
||||
## Deferred work and current priority
|
||||
|
||||
Until R4 closes, the following are not current blockers:
|
||||
The current non-UI priority order is:
|
||||
|
||||
1. keep E37–E40 status language methodologically honest;
|
||||
2. make the E41 predictor/evaluator boundary and E42 invariants mandatory
|
||||
regression checks;
|
||||
3. wire native pipeline telemetry into durable worker stages without inferring
|
||||
absent measurements;
|
||||
4. design exact-content references/deduplication from E44 before any storage
|
||||
migration;
|
||||
5. preserve E43 unchanged until the owner supplies the new capture.
|
||||
|
||||
The following remain outside the present stabilization increment:
|
||||
|
||||
- TEST007 qualification;
|
||||
- E36 second-source transfer;
|
||||
- a new physical route or camera capture;
|
||||
- E36 replay until an admitted source exists;
|
||||
- initiation of a new physical route/camera capture by an agent;
|
||||
- second K1 or changed mount generalization;
|
||||
- open-world and adversarial evaluation.
|
||||
|
||||
The existing E36 audit and its Ops card remain retained so transfer work can
|
||||
resume later without reconstructing history.
|
||||
The owner may provide a future recording under E43. Until it exists, current
|
||||
work neither depends on it nor simulates its result. The existing E36 audit and
|
||||
its Ops card remain retained so transfer work can resume without reconstructing
|
||||
history.
|
||||
|
||||
@@ -22,6 +22,14 @@ The existing Worker 006 containers were renamed in place to
|
||||
container identities were preserved, so this namespace migration did not restart the
|
||||
inference or perception runtimes.
|
||||
|
||||
The compute-side native pipeline contract is now implemented. It emits
|
||||
`missioncore.agent-pipeline-telemetry/v1` lifecycle documents through an injected
|
||||
transport, and the telemetry normalizer preserves source, method, stage and stage-state
|
||||
identities in the normalized series key. The E41 runner has exercised the JSONL
|
||||
evidence sink against the immutable predictor package. This proves producer and
|
||||
normalizer compatibility; it does not claim that the durable Worker 006 process has
|
||||
been wired to the MQTT sink or deployed with this code.
|
||||
|
||||
## Decision
|
||||
|
||||
Mission Core treats a compute worker as a configurable **local compute contour**, not as
|
||||
@@ -188,6 +196,7 @@ subscribe to `mission-core/v1/contours/+/agents/+/+`.
|
||||
- Mission Core code stays responsible for product semantics and stable API contracts,
|
||||
not OS-specific metric collection.
|
||||
- Host, container, network and inference telemetry is stored in the normalized path.
|
||||
Per-stage LAB processing telemetry remains unavailable until a laboratory worker
|
||||
publishes the native `runtime` and `pipeline` topic contracts; the UI must keep those
|
||||
stages explicitly unavailable rather than infer them from aggregate hardware load.
|
||||
- Per-stage LAB processing telemetry has a native producer contract and an admitted
|
||||
already-connected MQTT sink. It remains unavailable for a durable worker run until
|
||||
that worker actually injects the sink and publishes the `pipeline` topic; consumers
|
||||
must not infer stages from aggregate hardware load.
|
||||
|
||||
@@ -0,0 +1,89 @@
|
||||
# ADR 0032 — RAVNOVES00 methodology correction before transfer
|
||||
|
||||
Date: 2026-07-28
|
||||
|
||||
Status: accepted for implementation
|
||||
|
||||
## Context
|
||||
|
||||
E37–E40 produced useful, immutable source-scoped engineering evidence, but the
|
||||
architecture report was written before the source implementation and artifacts were
|
||||
audited together. E41 reproduced the exact data lineage and found four methodological
|
||||
limits:
|
||||
|
||||
1. E37 contains `484` engineering-reviewed labels and `2` human exception labels; it
|
||||
explicitly declares `independent_ground_truth=false`.
|
||||
2. The nominal `340` development / `146` validation split is not group-independent.
|
||||
Development and validation overlap by `14` exact frames, `38` track identities,
|
||||
`58` fixed 50-frame time blocks and `59` connected whole-track-or-scene groups.
|
||||
3. The E40 worker package co-located prediction inputs with the visible acceptance
|
||||
rows and split assignments.
|
||||
4. E38–E40 therefore measure performance against a historically evaluated,
|
||||
source-scoped engineering contract. They are not blind accuracy gates.
|
||||
|
||||
This does not invalidate their physical predictions, closed accounting or conservative
|
||||
authority. It changes the claim that those artifacts are allowed to support.
|
||||
|
||||
## Decision
|
||||
|
||||
1. Historical E37–E40 artifacts remain immutable evidence. Their metrics are retained,
|
||||
but “sealed”, “blind” and independent-accuracy language is removed from current
|
||||
status.
|
||||
2. The current `146`-item slice is named **historical evaluated visible validation**.
|
||||
It may compare immutable methods against the source-scoped engineering contract; it
|
||||
may not prove independent perception accuracy or cross-route generalization.
|
||||
3. Prediction and evaluation are physically separated. The E41 predictor package and
|
||||
result contain no acceptance rows, references, split assignments, severity or
|
||||
scoring state. The visible evaluator joins predictions to E37 only after inference.
|
||||
4. Runtime identity records the environment that actually executed a result. A local
|
||||
run records the repository lock, Python and NumPy identities and is not represented
|
||||
as a Worker 006 container run.
|
||||
5. Structural invariants are exercised independently of the visible labels. E42
|
||||
verifies predictor identity, row-order and chunk-boundary invariance plus PointSlab
|
||||
row-order invariance and rigid-coordinate equivariance.
|
||||
6. Native per-stage telemetry is a compute contract. Stages publish
|
||||
`started/completed/failed`, exact source/method/run identity, duration and accounting
|
||||
through an injected sink. A JSONL evidence sink and an already-connected MQTT sink
|
||||
are admitted; no network transport starts implicitly.
|
||||
7. Current storage work starts with content-addressed references and exact-content
|
||||
deduplication. E44 measures `1.680146×` amplification across the admitted E30–E40
|
||||
package/result roots. It does not authorize a storage-format or database migration.
|
||||
8. The later same-K1 transfer capture is preregistered before collection. E43 freezes
|
||||
minimum streams, route/control-bridge duration, connected scene/track/time
|
||||
partitioning, independent human review and truth reveal only after frozen
|
||||
prediction.
|
||||
9. Product windows, page anatomy and new design work are outside this stabilization
|
||||
decision and remain a separate owner review.
|
||||
|
||||
## Current immutable evidence
|
||||
|
||||
- methodology audit:
|
||||
`e41-methodology-audit-703dac176f20843ad6fbb9a1060b199942fba05e3bb34d3e2264ee87c7bfdd96`;
|
||||
- truth-free predictor package:
|
||||
`e41-predictor-package-fb35f42698013c63c6d417fd3b26986e9d3b89d12dce138bab18fe4203d593ad`;
|
||||
- truth-free prediction result:
|
||||
`e41-predictions-adcc1671de44646145a396d5815e522d6b4257d6ead9ace995761751655b3aac`;
|
||||
- visible engineering evaluation:
|
||||
`e41-visible-evaluation-57aa6c8569e8339630406e3c88cd539df13917fb08c1af57f0baae5cdd1069d2`;
|
||||
- metamorphic suite:
|
||||
`e42-metamorphic-suite-c358204e0b9e10eea5871e90b7c4fbe2c132f1f6d87074b2c5ffec18eaeebb40`;
|
||||
- preregistered future-capture protocol:
|
||||
`e43-future-capture-protocol-28f091b9648daffce988d44c183e21f56d77988061630de934f8003fb13701d8`;
|
||||
- data-amplification audit:
|
||||
`e44-data-amplification-89791978894ec785009e5b76c9325de4d9e910237af57476c9331eb47a8ccca4`.
|
||||
|
||||
## Consequences
|
||||
|
||||
- RAVNOVES00 remains the active engineering reference source, but R0 is reopened as an
|
||||
independent-truth and leakage-free-partition gate.
|
||||
- E40 does not pass R1: the visible evaluation is `84.2466%` for presence,
|
||||
`84.2466%` for geometry association and `94.5205%` for freshness, with `11`
|
||||
high-severity failures, complete accounting and zero false-free claims.
|
||||
- Further threshold or model tuning against the visible 146 items cannot close the
|
||||
blind gate. It may only create another explicitly visible engineering comparison.
|
||||
- Work that does not require a new recording can proceed now: methodology enforcement,
|
||||
predictor/evaluator separation, invariants, telemetry, exact-content deduplication
|
||||
design and preregistration.
|
||||
- A later new-route recording can close an independent transfer question only if the
|
||||
E43 protocol passes before labels are revealed. Until that capture exists,
|
||||
cross-route quality remains unknown rather than failed or inferred.
|
||||
Reference in New Issue
Block a user