SWI: Speaking with Intent in Large Language Models

2025-03-27

AI Summary

Large language models (LLMs) often lack explicit intent representation and high-level planning during reasoning, leading to suboptimal coherence, interpretability, and factual consistency. Method: We propose “Speaking with Intent” (SWI), the first approach to formally integrate cognitive-scientific notions of *intent* into LLM inference—requiring models to explicitly generate interpretable, high-level intents prior to multi-step reasoning and text generation, serving as anchoring guidance for subsequent steps. SWI is implemented via prompt engineering, requiring no fine-tuning or additional training and remaining compatible with mainstream LLM architectures. Contribution/Results: Extensive experiments demonstrate SWI’s significant gains over baselines—including Chain-of-Thought and Plan-and-Solve—across mathematical reasoning, question answering, and summarization tasks. Notably, it improves summary factuality and reduces hallucination. Human evaluation confirms that SWI-generated intents are coherent, effective, and highly interpretable, establishing a novel cognition-informed paradigm for reasoning enhancement.

Technology Category

• Cognitive Modeling & Cognitive Systems: Conceptual Inference and Reasoning
• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: (Large) Language Models

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Search Tool Learning with LLM: Teaching LLMs to invoke search and make use of retrieved information
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation

Abstract

Intent, typically clearly formulated and planned, functions as a cognitive framework for reasoning and problem-solving. This paper introduces the concept of Speaking with Intent (SWI) in large language models (LLMs), where the explicitly generated intent encapsulates the model's underlying intention and provides high-level planning to guide subsequent analysis and communication. By emulating deliberate and purposeful thoughts in the human mind, SWI is hypothesized to enhance the reasoning capabilities and generation quality of LLMs. Extensive experiments on mathematical reasoning benchmarks consistently demonstrate the superiority of Speaking with Intent over Baseline (i.e., generation without explicit intent). Moreover, SWI outperforms answer-trigger prompting methods Chain-of-Thought and Plan-and-Solve and maintains competitive performance with the strong method ARR (Analyzing, Retrieving, and Reasoning). Additionally, the effectiveness and generalizability of SWI are solidified on reasoning-intensive question answering (QA) and text summarization benchmarks, where SWI brings consistent improvement to the Baseline generation. In text summarization, SWI-generated summaries exhibit greater accuracy, conciseness, and factual correctness, with fewer hallucinations. Furthermore, human evaluations verify the coherence, effectiveness, and interpretability of the intent produced by SWI. This proof-of-concept study creates a novel avenue for enhancing LLMs' reasoning abilities with cognitive notions.

Problem

• Enhancing LLM reasoning via explicit intent generation
• Improving generation quality in QA and summarization
• Reducing hallucinations in text summarization outputs

Innovation

• Explicit intent generation guides LLM reasoning
• SWI enhances accuracy and reduces hallucinations
• Human-like cognitive framework improves generation quality