Recent work has shown that scaling large language models (LLMs) improves their alignment with human brain activity, yet it remains unclear what drives these gains and which representational properties are responsible. Although larger models often yield better task performance and brain alignment, they are increasingly difficult to analyze mechanistically. This raises a fundamental question: what is the minimal model capacity required to capture brain-relevant representations? To address this question, we systematically investigate how constraining model scale and numerical precision affects brain alignment. We compare full-precision LLMs, small language models (SLMs), and compressed variants (quantized and pruned) by predicting fMRI responses during naturalistic language comprehension. Across model families up to 14B parameters, we find that 3B SLMs achieve brain predictivity indistinguishable from larger LLMs, whereas 1B models degrade substantially, particularly in semantic language regions. Brain alignment is remarkably robust to compression: most quantization and pruning methods preserve neural predictivity, with GPTQ as a consistent exception. Linguistic probing reveals a dissociation between task performance and brain predictivity: compression degrades discourse, syntax, and morphology, yet brain predictivity remains largely unchanged. Overall, brain alignment saturates at modest model scales and is resilient to compression, challenging common assumptions about neural scaling and motivating compact models for brain-aligned language modeling.
Enabling effective brain-computer interfaces requires understanding how the human brain encodes stimuli across modalities such as visual, language (or text), etc. Brain encoding aims at constructing fMRI brain activity given a stimulus. There exists a plethora of neural encoding models which study brain encoding for single mode stimuli: visual (pretrained CNNs) or text (pretrained language models). Few recent papers have also obtained separate visual and text representation models and performed late-fusion using simple heuristics. However, previous work has failed to explore: (a) the effective…
Linguistics and Human Brain: A Perspective of Computational Neuroscience
Institute of AI and Robotics, College of Intelligent Robotics and Advanced Manufacturing, Fudan University（人工智能与机器人研究院，智能机器人与先进制造学院，复旦大学）；Department of Language Science and Technology, The Hong Kong Polytechnic University（语言科学与技术系，香港理工大学）；Department of Computing, The Hong Kong Polytechnic University（计算系，香港理工大学）
Left-right asymmetry in predicting brain activity from LLMs' representations emerges with their formal linguistic competence
Centre d’Analyse et de Mathématique Sociales CNRS, EHESS, Paris, France（分析与数学社会中心 CNRS，EHESS，巴黎，法国）；Cognitive Neuroimaging Unit CNRS, INSERM, CEA, Neurospin Center, 91191 Gif-sur-Yvette, France（认知神经成像单元 CNRS，INSERM，CEA，Neurospin中心，法国91191 Gif-sur-Yvette）
Far from the Shallow: Brain-Predictive Reasoning Embedding through Residual Disentanglement
Zuckerman Mind Brain Behavior Institute, Columbia University（扎克曼脑行为研究所，哥伦比亚大学）；Department of Electrical Engineering, Columbia University（电气工程系，哥伦比亚大学）；Department of Computer Science, Columbia University（计算机科学系，哥伦比亚大学）
Multilingual Brain Surgeon: Large Language Models Can be Compressed Leaving No Language Behind
X-LANCE Lab, Department of Computer Science and Engineering MoE Key Lab of Artificial Intelligence, SJTU AI Institute Shanghai Jiao Tong University, Shanghai, China（X-LANCE实验室，计算机科学与工程系，MoE人工智能关键实验室，上海交通大学人工智能研究院，上海，中国）；Suzhou Laboratory, Suzhou, China（苏州实验室，苏州，中国）
HIPPD: Brain-Inspired Hierarchical Information Processing for Personality Detection
School of Electronics and Computer Science University of Southampton（电子与计算机科学学院 英国南安普顿大学）；Department of Computational Biology Mohamed bin Zayed University of Artificial Intelligence（计算生物学系 摩洛哥本·扎耶德人工智能大学）
Brain-like Functional Organization within Large Language Models
School of Automation, Northwestern Polytechnical University（西北工业大学自动化学院）；School of Computing, University of Georgia（佐治亚大学计算学院）；School of Computer and Cyber Sciences, Augusta University（奥古斯塔大学计算机与网络科学学院）
Decoding the Multimodal Mind: Generalizable Brain-to-Text Translation via Multimodal Alignment and Adaptive Routing
State Key Laboratory of Multimodal Artificial Intelligence System, Institute of Automation, Chinese Academy of Sciences（多模态人工智能系统国家重点实验室，中国科学院自动化研究所）；School of Artificial Intelligence, University of Chinese Academy of Sciences（中国科学院大学人工智能学院）；Department of Computer Science, The University of Manchester（曼彻斯特大学计算机科学系）；Department of Language Science and Technology, Hong Kong Polytechnic University（香港理工大学语言科学与技术系）
Training-Driven Representational Geometry Modularization Predicts Brain Alignment in Language Models
School of Biomedical Engineering, Tsinghua University, Beijing, China（生物医学工程学院，清华大学，北京，中国）；Tsinghua Laboratory of Brain and Intelligence, Tsinghua University, Beijing, China（脑与智能实验室，清华大学，北京，中国）；School of Basic Medical Sciences, Tsinghua University, Beijing, China（基础医学学院，清华大学，北京，中国）；Department of Psychological and Cognitive Sciences, Tsinghua University, Beijing, China（心理学与认知科学系，清华大学，北京，中国）
Institute for Logic, Language and Computation（逻辑、语言与计算研究所）；University of Amsterdam（阿姆斯特丹大学）
In this paper, we define and apply representational stability analysis (ReStA), an intuitive way of analyzing neural language models. ReStA is a variant of the popular representational similarity analysis (RSA) in cognitive neuroscience. While RSA can be used to compare representations in models, model components, and human brains, ReStA compares instances of the same model, while systematically varying single model parameter. Using ReStA, we study four recent and successful neural language models, and evaluate how sensitive their internal representations are to the amount of prior context. Us…
Hierarchical Frequency Tagging Probe (HFTP): A Unified Approach to Investigate Syntactic Structure Representations in Large Language Models and the Human Brain
Peking University（北京大学）；Zhejiang University（浙江大学）；Beijing Language and Culture University（北京语言大学）；Beijing Institute for General Artificial Intelligence（北京通用人工智能研究院）
Inria（法国国家信息与自动化研究所）；Facebook AI Research（脸书人工智能研究院）；École normale supérieure（巴黎高等师范学院）；PSL University（巴黎文理研究大学）；CNRS（法国国家科学研究中心）
The activations of language transformers like GPT-2 have been shown to linearly map onto brain activity during speech comprehension. However, the nature of these activations remains largely unknown and presumably conflate distinct linguistic classes. Here, we propose a taxonomy to factorize the high-dimensional activations of language models into four combinatorial classes: lexical, compositional, syntactic, and semantic representations. We then introduce a statistical method to decompose, through the lens of GPT-2's activations, the brain activity of 345 subjects recorded with functional magn…
Instituto Nacional de Astrofísica, Óptica y Electrónica（国家天体物理、光学与电子研究所）
A Brain Computer Interface based on imagined words can decode the word a subject is thinking on through brain signals to control an external device. In order to build a fully asynchronous Brain Computer Interface based on imagined words in electroencephalogram signals as source, we need to solve the problem of detecting the onset of the imagined words. Although there has been some research in this field, the problem has not been fully solved. In this paper we present an approach to solve this problem by using values from statistics, information theory and chaos theory as features to correctly …
A Pre-trained Framework for Multilingual Brain Decoding Using Non-invasive Recordings
Shenzhen Institute of Advanced Technology（深圳先进技术研究院）；Chinese Academy of Sciences（中国科学院）；University of Chinese Academy of Sciences（中国科学院大学）；Hong Kong Baptist University（香港 Baptist大学）
The dispute of how the human brain represents conceptual knowledge has been argued in many scientific fields. Brain imaging studies have shown that the spatial patterns of neural activation in the brain are correlated with thinking about different semantic categories of words (for example, tools, animals, and buildings) or when viewing the related pictures. In this paper, we present a computational model that learns to predict the neural activation captured in functional magnetic resonance imaging (fMRI) data of test words. Unlike the models with hand-crafted features that have been used in th…
Distinct social-linguistic processing between humans and large audio-language models: Evidence from model-brain alignment
Department of Linguistics and Modern Languages, The Chinese University of Hong Kong（语言学与现代语言系，香港中文大学）；Brain and Mind Institute, The Chinese University of Hong Kong（脑与心智研究所，香港中文大学）
Transport of neural signals in the brain is challenging, owing to neural interference and neural noise. There is experimental evidence of multiplexing of sensory information across population of neurons, particularly in the vertebrate visual and olfactory systems. Recently, it has been discovered that in lateral intraparietal cortex of the brain, decision signals are multiplexed with decision-irrelevant visual signals. Furthermore, it is well known that several cortical neurons exhibit chaotic spiking patterns. Multiplexing of chaotic neural signals and their successful demultiplexing in the n…
University of Oxford（牛津大学）；Max Planck Institute for Psycholinguistics（马克斯·普朗克心理语言学研究所）；Radboud University（拉德堡德大学）
Recent artificial neural networks that process natural language achieve unprecedented performance in tasks requiring sentence-level understanding. As such, they could be interesting models of the integration of linguistic information in the human brain. We review works that compare these artificial language models with human brain activity and we assess the extent to which this approach has improved our understanding of the neural processes involved in natural language comprehension. Two main results emerge. First, the neural representation of word meaning aligns with the context-dependent, de…
Lost in Translation: The Algorithmic Gap Between LMs and the Brain
CHU Sainte Justine Research Center（圣贾斯汀医院研究中心）；Université de Montréal（蒙特利尔大学）；Mila（米拉研究所）；Ernst Strüngmann Institute for Neuroscience in Cooperation with Max Planck Society（恩斯特·施特龙格曼神经科学研究所（与马克斯·普朗克学会合作））
Convergent Representations of Linguistic Constructions in Human and Artificial Neural Systems
Department of English and American Studies, University Erlangen-Nuremberg（英语与美国研究系，埃尔朗根-纽伦堡大学）；Pattern Recognition Lab, University Erlangen-Nuremberg（模式识别实验室，埃尔朗根-纽伦堡大学）；Neuromodulation and Neuroprosthetics, University Hospital Mannheim, University Heidelberg（神经调控与神经假体，曼海姆大学医院，海德堡大学）；BGU Ludwigshafen, Germany（吕贝克大学吕贝克分校，德国）；Neuroscience Lab, University Hospital Erlangen（神经科学实验室，埃尔朗根大学医院）
Decoding Predictive Inference in Visual Language Processing via Spatiotemporal Neural Coherence
Department of Physics（物理系）；Carnegie-Mellon University（卡内基梅隆大学）；Department of Linguistics（语言学系）；University of Salzburg（萨尔茨堡大学）；Purdue University（普渡大学）；University of Alabama（阿拉巴马大学）；Department of Communicative Disorders（沟通障碍系）
The Temporal Structure of Language Processing in the Human Brain Corresponds to The Layered Hierarchy of Deep Language Models
Princeton University（普林斯顿大学）；Google Research（谷歌研究院）；Hebrew University of Jerusalem（希伯来大学）；University of California, Los Angeles（加利福尼亚大学洛杉矶分校）；New York University Grossman School of Medicine（纽约大学格罗斯曼医学院）
