← Back to reference

False Positive Analysis

Definition

A false positive is when the engine fires a deterministic rule on a non-fault event. We track:

  • False-positive rate = (FP / total non-fault observations).
  • FP burst rate = average FP count per cluster-hour.

Measurement

From 30 days of shadow-mode evaluation on production-equivalent clusters (anonymized):

Cluster tierFP rateFP per cluster-hour
64 GPU0.6 %0.16
1024 GPU0.9 %0.51
4096 GPU1.2 %1.42

Top Causes

  1. Threshold sensitivityR1 (Xid) fires on Xid 62 (training interrupts that are benign in some frameworks). Fix: replace ≥ 1 with = in allow-list; default to allow-list {48, 63, 64, 79, 94}.
  2. Counter resetsSBE>0 fired on counter wrap-around that lasted seconds. Fix: rate of change instead of absolute.
  3. NCCL log noiseNCCL async error rule fired on WARN lines that were recoverable. Fix: require ERROR level.

Drift Suppression

docs/operations/drift-suppression.md describes the routing of suspected FPs into a "shadow" bucket so users can inspect them.

How to Report a False Positive

Send the incident id + a short note to our support team. We push a rule adjustment within one release cycle.