skip to content
Kernel settings

Policy is explicit, not implicit

Thresholds decide when the kernel is allowed to stay quiet. They are shown here so a reader can disagree with them.

Demo modelSynthetic dataResearch preview

Inference backend

Local ONNX is the default path. The Hugging Face Inference Endpoint is optional and only used once it is configured and answering.

running on local onnx
runtime
endpoint url
endpoint healthnot provisioned
avg latency
p95 latency
error rate
windowno calls recorded
queueidle (max 2 in flight, 8 waiting)
failoverclosed
local fallbacksnone
last probenever

No endpoint is provisioned yet. Latency and error rate stay empty rather than showing a fabricated baseline — the local runtime's own timings are reported in the model version panel.

health trendp95 latency (left) · error rate and health (right)
Trend appears once the endpoint has answered a few calls.

Requests are queued at most 2 at a time with a 15s wait deadline and a 20s request deadline. Anything past those deadlines, or past the queue limit, is scored locally instead — never dropped, never left hanging.

Kernel thresholds

Preview build: adjustments are local to this session and are not persisted.

Surprise signal threshold0.30

below this, a typed signal is treated as ordinary variation

Unknown escalation threshold0.50

above this, the unknown channel forces an information request

Tail credibility floor0.25

scenarios below this are generated but not surfaced

Disagreement fragility threshold0.45

above this, model validity degrades to fragile

Admission control policy

  • Ignore — variation is inside the reference envelope.
  • Remember as exception — real but non-structural; stored in surprise memory.
  • Open new regime — the reference set no longer describes the system.
  • Request more information — the unknown channel dominates the signal.
  • Escalate for human review — impact is high and credibility is contested.

Provenance

zeno-kernel 0.4.0-previewsynthetic referencesdeterministic seed

All numbers in this preview come from deterministic synthetic reference models running in the browser. No trained forecast, live feed or proprietary dataset is used. Domain packs loaded: Earth, Finance, Critical.