Skip to content

Interpreting Certificates

PaulTiffany edited this page Nov 14, 2025 · 1 revision

Interpreting Certificates

Quick-Scan Infobox

Aspect Details
Target user Engineers and educators interpreting routing decisions
Core goal Map certificate fields to actionable interpretation
Time to first insight 5 minutes: read, compare to one run
Known limitation Ranges are typical, not absolute; tune to your data

Key Signals and How to Read Them

  • Utility (scalar): 0.5–0.9 typical. Lower values can indicate tight constraints or a novel/off‑manifold prompt. Compare relative changes across runs.
  • Boundary gap (preference separation): >0.15 clear preference; <0.05 ambiguous. See Math-Boundary-Gap-and-Entropy.
  • Entropy (decision spread): lower → concentrated preference; higher → diffuse confidence. Interpret alongside gap.
  • Shadow prices (constraint sensitivity): near 0 → not limiting; >0.5 → limiting utility. Focus remediation where largest.
  • Feasibility and slack: negative slack pinpoints the violated/active constraint; investigate policy or bounds first.
  • Drift (stability): small negative drift across iterations indicates stable improvement; large magnitude or positive values suggest instability.

A Compact Flow

graph TD
  A[Certificate] --> B{Gap > 0.15?}
  B -- Yes --> C[Stable choice]
  B -- No --> D[Ambiguous]
  A --> E{Max shadow > 0.5?}
  E -- Yes --> F[Constraint-limited]
  E -- No --> G[Capacity available]
  A --> H{Utility < 0.6?}
  H -- Yes --> I[Recalibrate / Review data]
  H -- No --> J[Proceed]
Loading

Caption: Decision aid for reading certificates. Alt text: Graph branching on gap, shadow prices, utility.


Typical Actions by Signal

  • High shadow price on one constraint (>0.5): review that bound; consider relaxing or revising policy.
  • Multiple high shadow prices: multiple bottlenecks; adjust one at a time and track utility.
  • Ambiguous boundary (gap <0.05): collect additional data near decision boundary or adjust temperature.
  • Low utility despite feasibility: revisit metric calibration and prompt family; examine off‑manifold indicators.
  • Increasing drift across trials: reduce trust region, check seeds, or pause batch changes.

Where the Math Lives

  • Boundary/entropy: see Math-Boundary-Gap-and-Entropy
  • Constraint duals/shadow prices: see Math-Constraints-Duality-and-Shadow-Prices
  • SPD metric and geometry: see Math-SPD-Metrics-and-Riemannian-Geometry
  • Coherence and density: see Math-Kernel-Density-and-Coherence

Practical Checklist

  • Read gap + entropy first: confident vs ambiguous.
  • Scan shadow prices: identify top constraint pressures.
  • Confirm feasibility + slack: any violations?
  • Track utility and drift across runs: trending in the right direction?
  • Write down a single action: change one variable at a time.

Clone this wiki locally