![[2602.22790] Natural Language Declarative Prompting (NLD-P): A Modular Governance Method for Prompt Design Under Model Drift](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.22790] Natural Language Declarative Prompting (NLD-P): A Modular Governance Method for Prompt Design Under Model Drift
Summary
The paper introduces Natural Language Declarative Prompting (NLD-P), a governance method for prompt design that addresses challenges posed by model drift in large language models (LLMs).
Why It Matters
As LLMs evolve, traditional prompt engineering methods become inadequate. NLD-P offers a structured approach to maintain control and interpretability, making it relevant for practitioners navigating the complexities of AI model governance.
Key Takeaways
- NLD-P is a modular governance framework for prompt design.
- It separates key components like provenance and evaluation for clarity.
- The method is designed for non-developers, enhancing accessibility.
- It addresses challenges posed by model drift in LLMs.
- Future research directions include empirical validation of NLD-P.
Computer Science > Computation and Language arXiv:2602.22790 (cs) [Submitted on 26 Feb 2026] Title:Natural Language Declarative Prompting (NLD-P): A Modular Governance Method for Prompt Design Under Model Drift Authors:Hyunwoo Kim, Hanau Yi, Jaehee Bae, Yumin Kim View a PDF of the paper titled Natural Language Declarative Prompting (NLD-P): A Modular Governance Method for Prompt Design Under Model Drift, by Hyunwoo Kim and 3 other authors View PDF HTML (experimental) Abstract:The rapid evolution of large language models (LLMs) has transformed prompt engineering from a localized craft into a systems-level governance challenge. As models scale and update across generations, prompt behavior becomes sensitive to shifts in instruction-following policies, alignment regimes, and decoding strategies, a phenomenon we characterize as GPT-scale model drift. Under such conditions, surface-level formatting conventions and ad hoc refinement are insufficient to ensure stable, interpretable control. This paper reconceptualizes Natural Language Declarative Prompting (NLD-P) as a declarative governance method rather than a rigid field template. NLD-P is formalized as a modular control abstraction that separates provenance, constraint logic, task content, and post-generation evaluation, encoded directly in natural language without reliance on external orchestration code. We define minimal compliance criteria, analyze model-dependent schema receptivity, and position NLD-P as an accessible gov...
