Skip to content

Assessment Feature

Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. SPDX-License-Identifier: MIT-0

This guide has moved. Extraction, confidence, and geometry documentation is now consolidated in Extraction & Confidence.

As of config v0.6, confidence scoring is an output of extraction (configured under extraction.confidence.*, with HITL under the top-level hitl.*). What moved:

  • Confidence assessment — the three modes (off / separate / integrated), the confidence model, in-shard vs standalone execution, and prompt placeholders.
  • Large-list batching (extraction.confidence.list_batch_size) — the replacement for the now-retired granular assessment; see Granular Assessment Retirement.
  • Confidence thresholds, UI color coding, and output format (explainability_info).

See Extraction & Confidence.