Section 013 · Chapter 2, From Tests to Release Evidence

Reproducibility: Logging the Right Things

Non-deterministic bugs are hard to debug unless Confidence Engineers capture the context around the failure.

reproducibilityconfidence engineerreproducibility logging right things

What to do

  1. Save the repo commit, branch, dependency lockfile, OS image, browser version, environment variables, test seed, exact command, failing output, screenshots, tool calls, files inspected, model version, prompt template, and diff.
  2. Define runnable checks that exercise reproducibility, confidence engineer, and reproducibility logging right things.
  3. Set acceptable outcomes and blocker failures for reproducibility, confidence engineer, and reproducibility logging right things before running the evaluation.

Evidence to preserve

  • Save the repo commit, branch, dependency lockfile, OS image, browser version, environment variables, test seed, exact command, failing output, screenshots, tool calls, files inspected, model version, prompt template, and diff.
  • Preserve the inputs, versions, configurations, raw outcomes, and results for reproducibility, confidence engineer, reproducibility logging right things needed to reproduce work on Reproducibility: Logging the Right Things.
  • Report results for reproducibility, confidence engineer, reproducibility logging right things by relevant slice, separate blocker failures from averages, state uncertainty and blind spots, and connect the result to a release decision.

Expert note

Expert logging distinguishes replay data from diagnosis data. Replay data tries to recreate conditions. Diagnosis data explains why the system behaved that way. Both should be privacy-aware, access-controlled, and tied to durable artifact IDs.

Continue the conversation

Apply this to your context.

Save your product context once, then open a focused conversation that combines it with this concept.

Cite this page

Jason Arbon. "Reproducibility: Logging the Right Things." Testing AI Knowledge Edition, section 13.

https://jarbon.ai/testing-ai/knowledge/ch013-reproducibility-logging-right-things.html

Shared across the Knowledge Edition

Adapt every concept to your world.

Describe your product, role, users, risks, constraints, or current quality problem. This stays in this browser until you choose to send it to ChatGPT.

Saved only in this browser.0 / 2400