Understanding AI Confidence Scores and Review Flags

Every mark that E-Porikkha awards comes with a confidence score — a number from 0 to 1 that reflects how certain the AI is that the awarded mark is correct. Understanding this score helps you focus your review time on the answers that actually need a second look.

The grading model compares the student's answer to the model answer across several dimensions: factual correctness, key terms present, logical structure, and marks-scheme alignment. The confidence score aggregates these signals.

A score above 0.85 means the AI is highly confident — typically a clear match or a clear miss. Scores between 0.5 and 0.85 indicate partial matches, ambiguous phrasing, or answers that are correct in spirit but worded differently from the model answer.

When a score falls below your review threshold (default: 0.70), the paper is marked Ready for Review and appears in your Papers queue with a badge count. Within the grading studio, the specific questions below threshold are highlighted in amber.

Flags are also added automatically when the AI detects a potential contradiction (e.g., the student says both X and not-X) or when an answer is very long relative to the question's mark allocation.

Go to Settings → AI Evaluation to change the review threshold. A higher threshold (e.g. 0.80) sends more papers to review — useful for high-stakes exams where you want to manually check every borderline case. A lower threshold (e.g. 0.60) lets more papers through automatically — appropriate for routine class tests.

You can set different thresholds per batch, overriding the school default.

In the grading studio, click the copilot icon on any flagged answer. The drawer opens a chat where you can ask the AI to justify its score, compare to a different part of the model answer, or propose a revised mark. You confirm or reject the proposal — the AI never changes marks without your explicit approval.