Know More, Know Clearer: A Meta-Cognitive Framework for Knowledge Augmentation in Large Language Models
This work addresses the issue of unreliable knowledge augmentation in LLMs for knowledge-intensive tasks, offering a novel approach to improve cognitive behaviors, though it is incremental in enhancing existing methods.
The paper tackles the problem of knowledge-confidence gaps in Large Language Models (LLMs) during knowledge augmentation, which cause overconfident errors or uncertain truths, and proposes a meta-cognitive framework that uses internal cognitive signals to partition knowledge and align certainty with accuracy, resulting in consistent performance improvements over strong baselines.
Knowledge augmentation has significantly enhanced the performance of Large Language Models (LLMs) in knowledge-intensive tasks. However, existing methods typically operate on the simplistic premise that model performance equates with internal knowledge, overlooking the knowledge-confidence gaps that lead to overconfident errors or uncertain truths. To bridge this gap, we propose a novel meta-cognitive framework for reliable knowledge augmentation via differentiated intervention and alignment. Our approach leverages internal cognitive signals to partition the knowledge space into mastered, confused, and missing regions, guiding targeted knowledge expansion. Furthermore, we introduce a cognitive consistency mechanism to synchronize subjective certainty with objective accuracy, ensuring calibrated knowledge boundaries. Extensive experiments demonstrate the our framework consistently outperforms strong baselines, validating its rationality in not only enhancing knowledge capabilities but also fostering cognitive behaviors that better distinguish knowns from unknowns.