Narration-of-Thought: Inference-Time Scaffolding for Defeasible Ethical Reasoning in Large Language Models
arXiv:2606.26366v1 Announce Type: new Abstract: Standard chain-of-thought on moral dilemmas exhibits two failure modes: stakeholder collapse (the trace names at most one party with a stake in the outcome) and uncertainty suppression (no explicit unknowns or hedges before committing to an action). We introduce narration-of-thought (NoT), a system prompt that structures chain-of-thought into five sections: protagonist, stakeholders, two-step consequences, uncertainty, then commitment. NoT adds no training, parameters, or fine-tuning. On 100 DailyDilemmas scenarios across four generators from thr
The proliferation of advanced LLMs necessitates robust ethical reasoning capabilities, especially as these models become more integrated into decision-making processes.
Improving ethical reasoning in LLMs addresses critical safety and alignment concerns, paving the way for more reliable and trustworthy AI systems in sensitive applications.
A new, scaffolded prompt engineering technique demonstrably enhances ethical reasoning by encouraging more comprehensive consideration of stakeholders and uncertainties, without requiring model retraining.
- · AI developers
- · Ethical AI researchers
- · Industries deploying AI in sensitive contexts
- · Developers relying solely on basic chain-of-thought
- · Users experiencing unpredictable AI ethical failures
Incorporation of 'narration-of-thought' methodologies becomes a standard practice in ethical AI development.
Increased public and regulatory trust in AI systems due to more transparent and defensible ethical decision-making processes.
Reduced likelihood of AI-induced moral hazards and externalities, accelerating AI adoption in highly regulated sectors.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.AI