Year: 2026
Type: conference-paper
Abstract: This study demonstrates an alignment of perword processing time in a popular state-space language model Mamba and human readers.In Mamba, the recurrent state transition at each layer conceptually takes some duration of time, the discretization timestep ∆ t , determined dynamically in response to the input.Using a naturalistic reading dataset, we show that the per-word timestep from Mamba is a significant predictor of human reading times, and remains significant even when known predictors such as GPT-2 surprisal are controlled for.We further suggest, through formal analysis of Mamba's architecture and internal dynamics, that Mamba can serve as a new, valuable lens to look at human real-time language processing with everupdated memory, because it allows us to look at how each module (layer) weighs short-and long-term information retention, and how noise may interact with dynamic, continuous memory representation.Code is available online. 1
Institutions: RIKEN, The University of Osaka, Tohoku University
Language: English
Subfield: Cognitive Neuroscience
Field: Neuroscience
Domain: Life Sciences
Open Access status: hybrid
Award: G1970772935
OpenAlex ID: W7166874311