Route M-Flight Stream Router
A deterministic browser-local research instrument.
6b471ddc- STREAMING
- SEGMENT_OPEN
- VALIDATED
- SEALED
- QUANTIZATION_MISMATCH
- FALLBACK
- REFUSED
Each control executes a declared browser-local transition. Repeated dual-outcome controls alternate deterministically.
NOT YET EXERCISEDNo segment seals before schema and exact digest validation.
NOT YET EXERCISEDEncoder and decoder must derive bit-identical probability integers.
NOT YET EXERCISEDOne segment failure does not corrupt prior sealed segments.
NOT YET EXERCISEDRaw fallback preserves recoverability.
No transition yet. Operate a declared control.
This transparent browser-local instrument teaches the authored state contract. Exercising an assertion demonstrates fixture behavior only; it is not evidence of production performance, compression ratio, model uplift, cognition, safety, or universal correctness.
Purpose
Show typed segments being validated and optionally entropy-coded during one generation flight.
Mechanism
A deterministic token-stream fixture emits section boundaries, route confidence, schema state, probability integers, and exact text. Completed segments seal independently with raw fallback.
Deterministic fixture
A structured research Return containing summary, claims, limitations, artifacts, and receipt fields plus one invalid segment and one probability-quantization drift.
The fixture must live in content/demos/d10-route-m-flight-stream/fixture.json, validate before the demo renders, and be resettable without a network call. A future live adapter may be added behind a visibly separate mode.
Required controls
- Play/pause stream
- Inspect token probabilities
- Mark cutline
- Validate segment
- Quantize CDF
- Seal M-flight layer
- Inject drift
- Use raw fallback
- Replay decode
Visible states
STREAMINGSEGMENT_OPENVALIDATEDSEALEDQUANTIZATION_MISMATCHFALLBACKREFUSED
Every state must have text, iconography, and a non-color-only distinction. State transitions are logged in an in-memory demo receipt visible in the inspector.
Core assertions
- No segment seals before schema and exact digest validation.
- Encoder and decoder must derive bit-identical probability integers.
- One segment failure does not corrupt prior sealed segments.
- Raw fallback preserves recoverability.
Layout
The page contains:
- a concise mechanism explanation;
- the interactive stage;
- a state and evidence inspector;
- a reset control;
- a “what this proves / what it does not prove” panel;
- links to the related paper and benchmark contract.
On narrow screens the inspector becomes a bottom sheet. The stage must remain usable at 360 CSS pixels.
Accessibility
- All operations are keyboard reachable.
- Pointer gestures have button or keyboard equivalents.
- Animated transitions honor
prefers-reduced-motion. - Focus remains visible and returns predictably after sheets close.
- Diagrams expose concise text alternatives.
- Status is announced through a polite live region only when the user initiates the transition.
Instrumentation
Record locally:
demo_id
fixture_version
action
prior_state
result_state
validation_result
timestamp_relative
Do not send telemetry. Provide “Export demo trace” as JSON for debugging and reproducibility.
Automated tests
- fixture schema validation;
- deterministic reset;
- every required control transition;
- every assertion above;
- invalid and stale fixture handling;
- keyboard path;
- 360 px and 1440 px screenshots;
- zero critical console errors;
- no request to a paid API in default mode.
Publication boundary
Probability data is deterministic fixture data unless a linked local-model run receipt is selected.
The interface must never convert illustrative values into a benchmark claim. A measured result appears only when linked to a versioned run receipt.