skip to content
Evidence

What is claimed, and what was actually measured

The honest summary first: the trained checkpoint fails the benchmark shipped with it. Everything below is either a link to that evidence or a statement of its limits.

claim NOT supported
Release
0.2.0
v0.2.0-trained
Checkpoint
c6a85149c344
synthetic pretrain only
Benchmark verdict
NOT supported
ac896a3a69dd
Real-data results
none
no validated external corpus yet

The claim under test

A learned representation of model-relative surprise contains information about future regime change and model failure that is not captured by predictive uncertainty alone.

  • — On the current synthetic benchmark the surprise index is inverted and fires on every clean episode, so it fails the degeneracy guards.
  • — A plain logistic model over the same signals is the strongest detector in the table. That is the number to beat, and it is not beaten.
  • — The uploaded-data path in Ingest and Signals therefore does not use the checkpoint. It uses stated reference predictors and a stated policy.

What would change the verdict

In order of expected effect.

  1. 1. Fix label/horizon alignment in the training objective; the inversion is consistent with a sign bug.
  2. 2. Train on a real corpus with generator-side labels replaced by measured reference degradation.
  3. 3. Re-run SurpriseBench with locked thresholds and the same controls, and publish the loss if it loses.
  4. 4. Only then let the checkpoint touch the scoring path used on user data.