Abstract
Memory formation is fundamental to intelligence, yet whether deep neural networks preserve identifiable memory traces analogous to biological memory units remains an open question. This work introduces a geometric framework to identify such "AI engrams" by formalizing the neuroscientific criteria of specificity, reactivation, sufficiency, and necessity into a constrained inverse problem.
We derive a closed-form estimator that isolates individual memory traces from globally entangled parameters, showing that this biologically-derived solution corresponds to a natural gradient update on the parameter manifold. AI engrams enable surgical manipulation of learned knowledge: any subset of memories can be composed or erased through linear arithmetic, without iterative optimization.
Experiments ranging from simple MLPs to LLMs demonstrate the causal validity and substantial scalability of AI engrams. Together, these results bridge theories of biological memory and artificial representation learning and offer geometric insight into how deep networks simultaneously support functional specificity within distributed storage.
Blogger's Review: This paper presents a novel geometric framework for understanding memory mechanisms in deep learning models, emphasizing the synergy between biology and AI. The effective identification and manipulation of memory traces could lead to more intelligent and adaptable AI systems, deepening our understanding of the essence of intelligence.