SpecMind: Cognitively Inspired, Interactive Multi-Turn Framework for Postcondition Inference

2026-02-24

AI Summary

Manually crafting program postconditions is time-consuming and error-prone, while existing single-prompt approaches leveraging large language models (LLMs) often yield suboptimal results. This work proposes SpecMind, a novel framework that treats the LLM as an interactive reasoner, iteratively refining postconditions through a cognitively inspired, multi-turn feedback mechanism that integrates both explicit and implicit correctness signals. Crucially, SpecMind autonomously determines when to terminate the refinement process based on its assessment of optimality. Experimental evaluation demonstrates that this approach significantly outperforms state-of-the-art methods in both the accuracy and completeness of the generated postconditions.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: Code Generation / Program Synthesis from Natural Language
• Cognitive Modeling & Cognitive Systems: Conceptual Inference and Reasoning

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation

Abstract

Specifications are vital for ensuring program correctness, yet writing them manually remains challenging and time-intensive. Recent large language model (LLM)-based methods have shown successes in generating specifications such as postconditions, but existing single-pass prompting often yields inaccurate results. In this paper, we present SpecMind, a novel framework for postcondition generation that treats LLMs as interactive and exploratory reasoners rather than one-shot generators. SpecMind employs feedback-driven multi-turn prompting approaches, enabling the model to iteratively refine candidate postconditions by incorporating implicit and explicit correctness feedback, while autonomously deciding when to stop. This process fosters deeper code comprehension and improves alignment with true program behavior via exploratory attempts. Our empirical evaluation shows that SpecMind significantly outperforms state-of-the-art approaches in both accuracy and completeness of generated postconditions.

Problem

• postcondition inference
• program specification
• large language models
• code correctness
• specification generation

Innovation

• multi-turn prompting
• postcondition inference
• feedback-driven refinement
• interactive reasoning
• large language models