In Praise of Stubbornness: The Case for Cognitive-Dissonance Aware Continual Update of Knowledge in LLMs

Simone Clemente, Zied Ben Houidi, Alexis Huet, Dario Rossi, Giulio Franzese, Pietro Michiardi

Keywords: LLM, knowledge editing, continual learning, Cognitive Science-Inspired AI, Human-inspired AI, Episodic Memory, Incremental Learning, Targeted Network Updates

TL;DR: The paper shows that updating LLMs with conflicting knowledge is destructive, even to unrelated prior knowledge, motivating the need for cognitive-dissonance awareness during continual update of knowledge in LLMs

Abstract:

Despite remarkable capabilities, large language models (LLMs) struggle to continually update their knowledge without catastrophic forgetting. In contrast, humans effortlessly integrate new information, detect conflicts with existing beliefs, and selectively update their mental models. This paper introduces a cognitive-inspired investigation paradigm to study knowledge updating in LLMs. We implement two key components inspired by human cognition: (1) *Dissonance and Familiarity Awareness*, analyzing model behavior to classify information as novel, familiar, or dissonant; and (2) *Targeted Network Updates*, which track neural activity to identify frequently used (*stubborn*) and rarely used (*plastic*) neurons.

Through carefully designed experiments in controlled settings, we uncover a number of empirical findings demonstrating the potential of this approach. First, dissonance detection is feasible using simple activation and gradient features, suggesting potential for cognitive-inspired training. Second, we find that non-dissonant updates largely preserve prior knowledge regardless of targeting strategy, revealing inherent robustness in LLM knowledge integration. Most critically, we discover that dissonant updates prove catastrophically destructive to the model's knowledge base, indiscriminately affecting even information unrelated to the current updates. This suggests fundamental limitations in how neural networks handle contradictions and motivates the need for new approaches to knowledge updating that better mirror human cognitive mechanisms.

Primary Area: foundation or frontier models, including LLMs

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.

Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2025/AuthorGuide.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Submission Number: 11976