Experience Is the Human LoRA: A Readout–Retention Theory of Selective Model Change
Zenodo (CERN European Organization for Nuclear Research) July 18, 2026 Yaoharee Lahtee
Experience is a finite event that transforms the observer, much like a low-rank update in a machine-learning system. From the inside, an experience feels like a bounded, embodied, emotionally weighted appearance directed at the world. From the outside, the same event selectively changes how the observer later perceives and acts. The argument uses a discrete mathematical framework where every accessible state is a finite difference on a finite graph, and time proceeds in steps. If an experience-driven update operates through a smaller active space within a larger organization, its change is low-rank. Durable learning comes from a retention mechanism, and cumulative change arises from many bounded updates. The theory connects to memory, neuroscience, and psychology, and treats mental suffering as a rigid but intelligible organization that has become harmful.