---
name: tai-ch137-testing-the-networks
description: 'Apply chapter 137 of Testing AI, Testing the Networks, as a workflow for evaluating AI and non-deterministic systems. Use for test planning, eval design, quality review, release evidence, examples, or coaching related to testing the networks.'
---

# Testing the Networks

Skill name: `tai-ch137-testing-the-networks`

Based on **Testing AI: Engineering Confidence in Non-Deterministic Systems** by **Jason Arbon**.

## Purpose

Network-aware confidence adds internal evidence to output testing without pretending activations
are perfect explanations.

## Use This Workflow

- Identify the AI behavior or release decision being evaluated.
- Define realistic cases, slices, unacceptable outcomes, and evidence needed for confidence.
- Choose measurements that match the risk: rubric scores, samples, intervals, traces, human review, deterministic checks, or production monitors.
- Report uncertainty, severe failures, and decision impact instead of only a pass/fail result.

## Key Guidance

Black-box output testing is still necessary. Users experience outputs, not activations. But for
high-risk AI systems, output checks alone can be too thin. Network-aware testing adds internal
evidence: what tokens were seen, where attention moved, which activations changed, and which
concept features appeared.

## Apply The Approach

Create representative cases, score them with explicit criteria, review severe failures separately, report uncertainty, and connect the evidence to a concrete decision.

## Deeper Guidance

Expert teams separate interpretability research from release evidence. Research may ask how a
circuit works. Testing asks whether an internal signal improves detection of regressions, risks,
coverage gaps, or unsafe behavior.