Module Glossary

Page Maps

graph LR
  family["Reproducible Research"]
  program["Deep Dive DVC"]
  section["Experiments Baselines Controlled Change"]
  page["Module Glossary"]
  capstone["Capstone evidence"]

  family --> program --> section --> page
  page -.applies in.-> capstone

flowchart LR
  orient["Orient on the page map"] --> read["Read the main claim and examples"]
  read --> inspect["Inspect the related code, proof, or capstone surface"]
  inspect --> verify["Run or review the verification path"]
  verify --> apply["Apply the idea back to the module and capstone"]

This glossary belongs to Module 06: Experiments, Baselines, and Controlled Change in Deep Dive DVC.

Use it to keep the module language stable while you move between the core lessons, the worked example, the exercises, and capstone review.

How to use this glossary

Read the directory index first. Return here when an experiment review, candidate comparison, baseline discussion, or promotion decision starts to feel vague.

The goal is not extra theory. The goal is shared language for exploring without damaging the baseline evidence story.

Terms in this directory

Term	Meaning in this directory
baseline	The recorded reference state used as the comparison anchor for candidates.
baseline authority	The degree to which baseline data, parameters, metrics, pipeline, and review evidence are trustworthy enough for comparison.
baseline boundary work	A change that redefines what the baseline means and should be reviewed before ordinary candidate comparison.
candidate	A provisional run being compared against a baseline or other candidates.
candidate intent	The short explanation of what a candidate is trying to learn or prove.
controlled change	A declared change with a reviewable purpose and evidence trail.
experiment scope	The set of changes included in one candidate run and the review question they answer.
experiment record	The DVC-preserved evidence for a candidate run, including parameter and metric differences.
experiment isolation	Keeping candidate runs separate from main Git history until a promotion decision is made.
comparison anchor	The baseline or prior state used as the reference for judging a candidate.
selection	The review decision to keep, discard, or promote a candidate based on evidence and intent.
metric tradeoff	A result where one metric improves while another worsens, requiring review judgment.
promotion	The deliberate decision to make a candidate part of the main state story.
apply	Bringing a candidate's state into the workspace for inspection before any promotion commit.
discard	Removing a candidate from active consideration because it is not useful, valid, comparable, or worth keeping.
workspace discipline	Keeping local files understandable while reviewing, applying, or discarding candidates.
lineage	The evidence trail that explains where a result came from and how it relates to the baseline.
local folklore	Unrecorded knowledge about candidate runs that depends on memory instead of inspectable evidence.
promotion note	A review statement explaining the control change, metric tradeoff, baseline comparison, and reason for promotion.
release objective	The decision criterion that tells reviewers which tradeoffs matter for promotion.

Stable review questions

Use these questions when the module feels abstract:

• What baseline is this candidate comparing against?
• What is the candidate trying to learn?
• Which declared control changed?
• Did the metric definition and population stay comparable?
• What tradeoff did the candidate create?
• Does the candidate deserve promotion, discard, or more bounded review?
• What changed after applying the candidate to the workspace?
• What should the promotion note say two years from now?