Bridging Brains and Models: MoE-Based Functional Lesions for Simulating and Rehabilitating Aphasia

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Wang, Yifan, Sun, Jingyuan, Zheng, Jichen, Zhang, Yunhao, Ye, Chunyu, Li, Jixing, Zong, Chengqing, Wang, Shaonan
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866912524804292608
author Wang, Yifan
Sun, Jingyuan
Zheng, Jichen
Zhang, Yunhao
Ye, Chunyu
Li, Jixing
Zong, Chengqing
Wang, Shaonan
author_facet Wang, Yifan
Sun, Jingyuan
Zheng, Jichen
Zhang, Yunhao
Ye, Chunyu
Li, Jixing
Zong, Chengqing
Wang, Shaonan
contents The striking alignment between large language models (LLMs) and human brain activity positions them as powerful models of healthy cognition. This parallel raises a fundamental question: if LLMs can model the intact brain, can we lesion them to simulate the linguistic deficits of the injured brain? In this work, we introduce a methodology to model aphasia - a complex language disorder caused by neural injury - by selectively disabling components in a modular Mixture-of-Experts (MoE) language model. We simulate distinct aphasia subtypes, validate their linguistic outputs against real patient speech, and then investigate functional recovery by retraining the model's remaining healthy experts. Our results demonstrate that lesioning functionally-specialized experts for syntax or semantics induces distinct impairments that closely resemble Broca's and Wernicke's aphasia, respectively. Crucially, we show that freezing the damaged experts and retraining the intact ones on conversational data restores significant linguistic function, demonstrating a computational analogue for rehabilitation. These findings establish modular LLMs as a powerful and clinically-relevant potential framework for modeling the mechanisms of language disorders and for computationally exploring novel pathways for therapy.
format Preprint
id arxiv_https___arxiv_org_abs_2508_04749
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Bridging Brains and Models: MoE-Based Functional Lesions for Simulating and Rehabilitating Aphasia
Wang, Yifan
Sun, Jingyuan
Zheng, Jichen
Zhang, Yunhao
Ye, Chunyu
Li, Jixing
Zong, Chengqing
Wang, Shaonan
Neurons and Cognition
The striking alignment between large language models (LLMs) and human brain activity positions them as powerful models of healthy cognition. This parallel raises a fundamental question: if LLMs can model the intact brain, can we lesion them to simulate the linguistic deficits of the injured brain? In this work, we introduce a methodology to model aphasia - a complex language disorder caused by neural injury - by selectively disabling components in a modular Mixture-of-Experts (MoE) language model. We simulate distinct aphasia subtypes, validate their linguistic outputs against real patient speech, and then investigate functional recovery by retraining the model's remaining healthy experts. Our results demonstrate that lesioning functionally-specialized experts for syntax or semantics induces distinct impairments that closely resemble Broca's and Wernicke's aphasia, respectively. Crucially, we show that freezing the damaged experts and retraining the intact ones on conversational data restores significant linguistic function, demonstrating a computational analogue for rehabilitation. These findings establish modular LLMs as a powerful and clinically-relevant potential framework for modeling the mechanisms of language disorders and for computationally exploring novel pathways for therapy.
title Bridging Brains and Models: MoE-Based Functional Lesions for Simulating and Rehabilitating Aphasia
topic Neurons and Cognition
url https://arxiv.org/abs/2508.04749