What question did this study set out to answer?

The research aims to preserve multidimensional expert judgment in LLM outputs using layered semantic constraints.

May 24, 2026Open Access

Semantic Drafting: Preserving Multidimensional Expert Judgment in LLM Output through Layered Constraints

Key Points

The research aims to preserve multidimensional expert judgment in LLM outputs using layered semantic constraints.
Proposed the Semantic Drafting framework with layered reasoning levels (L1-in to L1-out).
Conducted prompt experiments and evaluations in landscape-related tasks to assess LLM performance.
Identified and addressed failure modes through the introduction of Constraint Quality metrics and a management cycle.
Identified failure modes including bleaching, drift, forgetting, and structural collapse in LLM outputs.
Introduced L4 Constraint Quality for measuring and improving constraint effectiveness.
Demonstrated that structured reasoning improves the observability and quality of AI-assisted judgment.

Abstract

This paper proposes Semantic Drafting (SD), a framework for preserving multidimensional expert judgment in LLM outputs through layered semantic constraints. SD organizes reasoning through L1-in, L4, L3, L2, and L1-out, and examines how upper-level constraints propagate into lower-layer generation. Through SD prompt experiments and practitioner evaluation in landscape-related tasks, the paper identifies failure modes such as bleaching, drift, forgetting, and structural collapse. It introduces L4 Constraint Quality (LCQ), Authority Check, and an SD quality management cycle to diagnose and correct failures in constraint propagation. SD is positioned not as a replacement for expert judgment, but as a quality-managed reasoning design framework for making AI-assisted judgment more observable and improvable.

Read Full Paperexternally

Ask AI

Helpful

Bookmark

View Full Paper