Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

From "Thinking" to "Justifying": Aligning High-Stakes Explainability with Professional Communication Standards

Chen Qian, Yimeng Wang, Yu Chen, Lingfei Wu, Andreas Stathopoulos

TL;DR

This work tackles trustworthy explainability in high-stakes AI by shifting from tracing internal reasoning to presenting a conclusion-first justification. It introduces the Structured Explainability Framework (SEF), which operationalizes CREAC and BLUF conventions via six metrics that capture output plausibility and faithfulness. Empirical results across four Yes/No tasks and four 12–14B models show strong correlations between the metrics and correctness and demonstrate that SEF achieves 83.9% accuracy, outperforming Chain-of-Thought baselines by about 5 points. The findings suggest that structured, verifiable explanations can enhance reliability and accountability in domain-critical AI, while highlighting the need for human oversight and careful handling of potential persuasive inconsistencies.

Abstract

Explainable AI (XAI) in high-stakes domains should help stakeholders trust and verify system outputs. Yet Chain-of-Thought methods reason before concluding, and logical gaps or hallucinations can yield conclusions that do not reliably align with their rationale. Thus, we propose "Result -> Justify", which constrains the output communication to present a conclusion before its structured justification. We introduce SEF (Structured Explainability Framework), operationalizing professional conventions (e.g., CREAC, BLUF) via six metrics for structure and grounding. Experiments across four tasks in three domains validate this approach: all six metrics correlate with correctness (r=0.20-0.42; p<0.001), and SEF achieves 83.9% accuracy (+5.3 over CoT). These results suggest structured justification can improve verifiability and may also improve reliability.

From "Thinking" to "Justifying": Aligning High-Stakes Explainability with Professional Communication Standards

TL;DR

This work tackles trustworthy explainability in high-stakes AI by shifting from tracing internal reasoning to presenting a conclusion-first justification. It introduces the Structured Explainability Framework (SEF), which operationalizes CREAC and BLUF conventions via six metrics that capture output plausibility and faithfulness. Empirical results across four Yes/No tasks and four 12–14B models show strong correlations between the metrics and correctness and demonstrate that SEF achieves 83.9% accuracy, outperforming Chain-of-Thought baselines by about 5 points. The findings suggest that structured, verifiable explanations can enhance reliability and accountability in domain-critical AI, while highlighting the need for human oversight and careful handling of potential persuasive inconsistencies.

Abstract

Explainable AI (XAI) in high-stakes domains should help stakeholders trust and verify system outputs. Yet Chain-of-Thought methods reason before concluding, and logical gaps or hallucinations can yield conclusions that do not reliably align with their rationale. Thus, we propose "Result -> Justify", which constrains the output communication to present a conclusion before its structured justification. We introduce SEF (Structured Explainability Framework), operationalizing professional conventions (e.g., CREAC, BLUF) via six metrics for structure and grounding. Experiments across four tasks in three domains validate this approach: all six metrics correlate with correctness (r=0.20-0.42; p<0.001), and SEF achieves 83.9% accuracy (+5.3 over CoT). These results suggest structured justification can improve verifiability and may also improve reliability.
Paper Structure (56 sections, 1 figure, 5 tables)

This paper contains 56 sections, 1 figure, 5 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Process-Oriented Reasoning.
  4. Explainability Evaluation.
  5. Domain-Specific NLP.
  6. Sef Framework
  7. Six Justification Metrics.
  8. Metric-informed prompting.
  9. Experiments
  10. Setup
  11. Models.
  12. Tasks.
  13. Baselines.
  14. Ablation Protocol.
  15. Metric-Accuracy Correlation
  16. ...and 41 more sections

Figures (1)

  • Figure 1: Sef: Structured Explainability via CREAC/BLUF Conventions.Left: Paradigm comparison showing CoT ("Reasoning $\rightarrow$ Result") vs. Sef ("Result $\rightarrow$ Justify"). CoT traces can be hard to verify (Direct 80.9% $>$ CoT 78.6%); Sef achieves 83.9%. Right: PubMedQA example demonstrating the 4-step prompt (Commit, Ground, Defend, Conclude) with metric mappings.