Simultaneously uncovering the patterns of brain regions involved in different story reading subprocesses

Stacked regressions and structured variance partitioning for interpretable brain maps

Relating brain activity associated with a complex stimulus to different properties of that stimulus is a powerful approach for constructing functional brain maps. However, when stimuli are naturalistic, their properties are often correlated (e.g., visual and semantic features of natural images, or different layers of a convolutional neural network that are used as features of images). Correlated properties can act as confounders for each other and complicate the interpretability of brain maps, and can impact the robustness of statistical estimators. Here, we present an approach for brain mapping based on two proposed methods: stacking different encoding models and structured variance partitioning. Our stacking algorithm combines encoding models that each uses as input a feature space that describes a different stimulus attribute. The algorithm learns to predict the activity of a voxel as a linear combination of the outputs of different encoding models. We show that the resulting combined model can predict held-out brain activity better or at least as well as the individual encoding models. Further, the weights of the linear combination are readily interpretable; they show the importance of each feature space for predicting a voxel. We then build on our stacking models to introduce structured variance partitioning, a new type of variance partitioning that takes into account the known relationships between features. Our approach constrains the size of the hypothesis space and allows us to ask targeted questions about the similarity between feature spaces and brain regions even in the presence of correlations between the feature spaces. We validate our approach in simulation, showcase its brain mapping potential on fMRI data, and release a Python package. Our methods can be useful for researchers interested in aligning brain activity with different layers of a neural network, or with other types of correlated feature spaces.

The Cerebellum Is Sensitive to the Lexical Properties of Words During Spoken Language Comprehension

Over the past few decades, research into the function of the cerebellum has expanded far beyond the motor domain. A growing number of studies are probing the role of specific cerebellar subregions, such as Crus I and Crus II, in higher-order cognitive functions including receptive language processing. In the current fMRI study, we show evidence for the cerebellum's sensitivity to variation in two well-studied psycholinguistic properties of words-lexical frequency and phonological neighborhood density-during passive, continuous listening of a podcast. To determine whether, and how, activity in the cerebellum correlates with these lexical properties, we modeled each word separately using an amplitude-modulated regressor, time-locked to the onset of each word. At the group level, significant effects of both lexical properties landed in expected cerebellar subregions: Crus I and Crus II. The BOLD signal correlated with variation in each lexical property, consistent with both language-specific and domain-general mechanisms. Activation patterns at the individual level also showed that effects of phonological neighborhood and lexical frequency landed in Crus I and Crus II as the most probable sites, though there was activation seen in other lobules (especially for frequency). Although the exact cerebellar mechanisms used during speech and language processing are not yet evident, these findings highlight the cerebellum's role in word-level processing during continuous listening.

Language in Brains, Minds, and Machines

It has long been argued that only humans could produce and understand language. But now, for the first time, artificial language models (LMs) achieve this feat. Here we survey the new purchase LMs are providing on the question of how language is implemented in the brain. We discuss why, a priori, LMs might be expected to share similarities with the human language system. We then summarize evidence that LMs represent linguistic information similarly enough to humans to enable relatively accurate brain encoding and decoding during language processing. Finally, we examine which LM properties-their architecture, task performance, or training-are critical for capturing human neural responses to language and review studies using LMs as in silico model organisms for testing hypotheses about language. These ongoing investigations bring us closer to understanding the representations and processes that underlie our ability to comprehend sentences and express thoughts in language.

Semantic encoding during language comprehension at single-cell resolution

From sequences of speech sounds1,2 or letters3, humans can extract rich and nuanced meaning through language. This capacity is essential for human communication. Yet, despite a growing understanding of the brain areas that support linguistic and semantic processing4-12, the derivation of linguistic meaning in neural tissue at the cellular level and over the timescale of action potentials remains largely unknown. Here we recorded from single cells in the left language-dominant prefrontal cortex as participants listened to semantically diverse sentences and naturalistic stories. By tracking their activities during natural speech processing, we discover a fine-scale cortical representation of semantic information by individual neurons. These neurons responded selectively to specific word meanings and reliably distinguished words from nonwords. Moreover, rather than responding to the words as fixed memory representations, their activities were highly dynamic, reflecting the words' meanings based on their specific sentence contexts and independent of their phonetic form. Collectively, we show how these cell ensembles accurately predicted the broad semantic categories of the words as they were heard in real time during speech and how they tracked the sentences in which they appeared. We also show how they encoded the hierarchical structure of these meaning representations and how these representations mapped onto the cell population. Together, these findings reveal a finely detailed cortical organization of semantic representations at the neuron scale in humans and begin to illuminate the cellular-level processing of meaning during language comprehension.

Shared functional specialization in transformer-based language models and the human brain

When processing language, the brain is thought to deploy specialized computations to construct meaning from complex linguistic structures. Recently, artificial neural networks based on the Transformer architecture have revolutionized the field of natural language processing. Transformers integrate contextual information across words via structured circuit computations. Prior work has focused on the internal representations ("embeddings") generated by these circuits. In this paper, we instead analyze the circuit computations directly: we deconstruct these computations into the functionally-specialized "transformations" that integrate contextual information across words. Using functional MRI data acquired while participants listened to naturalistic stories, we first verify that the transformations account for considerable variance in brain activity across the cortical language network. We then demonstrate that the emergent computations performed by individual, functionally-specialized "attention heads" differentially predict brain activity in specific cortical regions. These heads fall along gradients corresponding to different layers and context lengths in a low-dimensional cortical space.

Reward-related regions play a role in natural story comprehension

The reward system was shown to be involved in a wide array of processes. Nevertheless, the exploration of the involvement of the reward system during language processing has not yet been directly tested. We investigated the role of reward-processing regions while listening to a natural story. We utilized a published dataset in which half of the participants listened to a natural story and the others listened to a scrambled version of it to compare the functional MRI signals in the reward system between these conditions and discovered a distinct pattern between conditions. This suggests that the reward system is activated during the comprehension of natural stories. We also show evidence that the fMRI signals in reward-related areas might potentially correlate with the predictability level of processed sentences. Further research is needed to determine the nature of the involvement and the way the activity interacts with various aspects of the sentences.

Computational Language Modeling and the Promise of In Silico Experimentation

Language neuroscience currently relies on two major experimental paradigms: controlled experiments using carefully hand-designed stimuli, and natural stimulus experiments. These approaches have complementary advantages which allow them to address distinct aspects of the neurobiology of language, but each approach also comes with drawbacks. Here we discuss a third paradigm-in silico experimentation using deep learning-based encoding models-that has been enabled by recent advances in cognitive computational neuroscience. This paradigm promises to combine the interpretability of controlled experiments with the generalizability and broad scope of natural stimulus experiments. We show four examples of simulating language neuroscience experiments in silico and then discuss both the advantages and caveats of this approach.

The neural and cognitive basis of expository text comprehension

As science and technology rapidly progress, it becomes increasingly important to understand how individuals comprehend expository technical texts that explain these advances. This study examined differences in individual readers' technical comprehension performance and differences among texts, using functional brain imaging to measure regional brain activity while students read passages on technical topics and then took a comprehension test. Better comprehension of the technical passages was related to higher activation in regions of the left inferior frontal gyrus, left superior parietal lobe, bilateral dorsolateral prefrontal cortex, and bilateral hippocampus. These areas are associated with the construction of a mental model of the passage and with the integration of new and prior knowledge in memory. Poorer comprehension of the passages was related to greater activation of the ventromedial prefrontal cortex and the precuneus, areas involved in autobiographical and episodic memory retrieval. More comprehensible passages elicited more brain activation associated with establishing links among different types of information in the text and activation associated with establishing conceptual coherence within the text representation. These findings converge with previous behavioral research in their implications for teaching technical learners to become better comprehenders and for improving the structure of instructional texts, to facilitate scientific and technological comprehension.

The cortical representation of language timescales is shared between reading and listening

Language comprehension involves integrating low-level sensory inputs into a hierarchy of increasingly high-level features. Prior work studied brain representations of different levels of the language hierarchy, but has not determined whether these brain representations are shared between written and spoken language. To address this issue, we analyze fMRI BOLD data that were recorded while participants read and listened to the same narratives in each modality. Levels of the language hierarchy are operationalized as timescales, where each timescale refers to a set of spectral components of a language stimulus. Voxelwise encoding models are used to determine where different timescales are represented across the cerebral cortex, for each modality separately. These models reveal that between the two modalities timescale representations are organized similarly across the cortical surface. Our results suggest that, after low-level sensory processing, language integration proceeds similarly regardless of stimulus modality.

Driving and suppressing the human language network using large language models

Transformer models such as GPT generate human-like language and are predictive of human brain responses to language. Here, using functional-MRI-measured brain responses to 1,000 diverse sentences, we first show that a GPT-based encoding model can predict the magnitude of the brain response associated with each sentence. We then use the model to identify new sentences that are predicted to drive or suppress responses in the human language network. We show that these model-selected novel sentences indeed strongly drive and suppress the activity of human language areas in new individuals. A systematic analysis of the model-selected sentences reveals that surprisal and well-formedness of linguistic input are key determinants of response strength in the language network. These results establish the ability of neural network models to not only mimic human language but also non-invasively control neural activity in higher-level cortical areas, such as the language network.

Information-Restricted Neural Language Models Reveal Different Brain Regions' Sensitivity to Semantics, Syntax, and Context

A fundamental question in neurolinguistics concerns the brain regions involved in syntactic and semantic processing during speech comprehension, both at the lexical (word processing) and supra-lexical levels (sentence and discourse processing). To what extent are these regions separated or intertwined? To address this question, we introduce a novel approach exploiting neural language models to generate high-dimensional feature sets that separately encode semantic and syntactic information. More precisely, we train a lexical language model, GloVe, and a supra-lexical language model, GPT-2, on a text corpus from which we selectively removed either syntactic or semantic information. We then assess to what extent the features derived from these information-restricted models are still able to predict the fMRI time courses of humans listening to naturalistic text. Furthermore, to determine the windows of integration of brain regions involved in supra-lexical processing, we manipulate the size of contextual information provided to GPT-2. The analyses show that, while most brain regions involved in language comprehension are sensitive to both syntactic and semantic features, the relative magnitudes of these effects vary across these regions. Moreover, regions that are best fitted by semantic or syntactic features are more spatially dissociated in the left hemisphere than in the right one, and the right hemisphere shows sensitivity to longer contexts than the left. The novelty of our approach lies in the ability to control for the information encoded in the models' embeddings by manipulating the training set. These "information-restricted" models complement previous studies that used language models to probe the neural bases of language, and shed new light on its spatial organization.

The Cortical Representation of Language Timescales is Shared between Reading and Listening

Language comprehension involves integrating low-level sensory inputs into a hierarchy of increasingly high-level features. Prior work studied brain representations of different levels of the language hierarchy, but has not determined whether these brain representations are shared between written and spoken language. To address this issue, we analyzed fMRI BOLD data recorded while participants read and listened to the same narratives in each modality. Levels of the language hierarchy were operationalized as timescales, where each timescale refers to a set of spectral components of a language stimulus. Voxelwise encoding models were used to determine where different timescales are represented across the cerebral cortex, for each modality separately. These models reveal that between the two modalities timescale representations are organized similarly across the cortical surface. Our results suggest that, after low-level sensory processing, language integration proceeds similarly regardless of stimulus modality.

Introducing MEG-MASC a high-quality magneto-encephalography dataset for evaluating natural speech processing

The "MEG-MASC" dataset provides a curated set of raw magnetoencephalography (MEG) recordings of 27 English speakers who listened to two hours of naturalistic stories. Each participant performed two identical sessions, involving listening to four fictional stories from the Manually Annotated Sub-Corpus (MASC) intermixed with random word lists and comprehension questions. We time-stamp the onset and offset of each word and phoneme in the metadata of the recording, and organize the dataset according to the 'Brain Imaging Data Structure' (BIDS). This data collection provides a suitable benchmark to large-scale encoding and decoding analyses of temporally-resolved brain responses to speech. We provide the Python code to replicate several validations analyses of the MEG evoked responses such as the temporal decoding of phonetic features and word frequency. All code and MEG, audio and text data are publicly available to keep with best practices in transparent and reproducible research.

Brain encoding models based on multimodal transformers can transfer across language and vision

Encoding models have been used to assess how the human brain represents concepts in language and vision. While language and vision rely on similar concept representations, current encoding models are typically trained and tested on brain responses to each modality in isolation. Recent advances in multimodal pretraining have produced transformers that can extract aligned representations of concepts in language and vision. In this work, we used representations from multimodal transformers to train encoding models that can transfer across fMRI responses to stories and movies. We found that encoding models trained on brain responses to one modality can successfully predict brain responses to the other modality, particularly in cortical regions that represent conceptual meaning. Further analysis of these encoding models revealed shared semantic dimensions that underlie concept representations in language and vision. Comparing encoding models trained using representations from multimodal and unimodal transformers, we found that multimodal transformers learn more aligned representations of concepts in language and vision. Our results demonstrate how multimodal transformers can provide insights into the brain's capacity for multimodal processing.

Modeling Brain Representations of Words' Concreteness in Context Using GPT-2 and Human Ratings

The meaning of most words in language depends on their context. Understanding how the human brain extracts contextualized meaning, and identifying where in the brain this takes place, remain important scientific challenges. But technological and computational advances in neuroscience and artificial intelligence now provide unprecedented opportunities to study the human brain in action as language is read and understood. Recent contextualized language models seem to be able to capture homonymic meaning variation ("bat", in a baseball vs. a vampire context), as well as more nuanced differences of meaning-for example, polysemous words such as "book", which can be interpreted in distinct but related senses ("explain a book", information, vs. "open a book", object) whose differences are fine-grained. We study these subtle differences in lexical meaning along the concrete/abstract dimension, as they are triggered by verb-noun semantic composition. We analyze functional magnetic resonance imaging (fMRI) activations elicited by Italian verb phrases containing nouns whose interpretation is affected by the verb to different degrees. By using a contextualized language model and human concreteness ratings, we shed light on where in the brain such fine-grained meaning variation takes place and how it is coded. Our results show that phrase concreteness judgments and the contextualized model can predict BOLD activation associated with semantic composition within the language network. Importantly, representations derived from a complex, nonlinear composition process consistently outperform simpler composition approaches. This is compatible with a holistic view of semantic composition in the brain, where semantic representations are modified by the process of composition itself. When looking at individual brain areas, we find that encoding performance is statistically significant, although with differing patterns of results, suggesting differential involvement, in the posterior superior temporal sulcus, inferior frontal gyrus and anterior temporal lobe, and in motor areas previously associated with processing of concreteness/abstractness.

Driving and suppressing the human language network using large language models

Transformer models such as GPT generate human-like language and are highly predictive of human brain responses to language. Here, using fMRI-measured brain responses to 1,000 diverse sentences, we first show that a GPT-based encoding model can predict the magnitude of brain response associated with each sentence. Then, we use the model to identify new sentences that are predicted to drive or suppress responses in the human language network. We show that these model-selected novel sentences indeed strongly drive and suppress activity of human language areas in new individuals. A systematic analysis of the model-selected sentences reveals that surprisal and well-formedness of linguistic input are key determinants of response strength in the language network. These results establish the ability of neural network models to not only mimic human language but also noninvasively control neural activity in higher-level cortical areas, like the language network.

A common neural code for meaning in discourse production and comprehension

How does the brain code the meanings conveyed by language? Neuroimaging studies have investigated this by linking neural activity patterns during discourse comprehension to semantic models of language content. Here, we applied this approach to the production of discourse for the first time. Participants underwent fMRI while producing and listening to discourse on a range of topics. We used a distributional semantic model to quantify the similarity between different speech passages and identified where similarity in neural activity was predicted by semantic similarity. When people produced discourse, speech on similar topics elicited similar activation patterns in a widely distributed and bilateral brain network. This network was overlapping with, but more extensive than, the regions that showed similarity effects during comprehension. Critically, cross-task neural similarities between comprehension and production were also predicted by similarities in semantic content. This result suggests that discourse semantics engages a common neural code that is shared between comprehension and production. Effects of semantic similarity were bilateral in all three RSA analyses, even while univariate activation contrasts in the same data indicated left-lateralised BOLD responses. This indicates that right-hemisphere regions encode semantic properties even when they are not activated above baseline. We suggest that right-hemisphere regions play a supporting role in processing the meaning of discourse during both comprehension and production.

A natural language fMRI dataset for voxelwise encoding models

Speech comprehension is a complex process that draws on humans' abilities to extract lexical information, parse syntax, and form semantic understanding. These sub-processes have traditionally been studied using separate neuroimaging experiments that attempt to isolate specific effects of interest. More recently it has become possible to study all stages of language comprehension in a single neuroimaging experiment using narrative natural language stimuli. The resulting data are richly varied at every level, enabling analyses that can probe everything from spectral representations to high-level representations of semantic meaning. We provide a dataset containing BOLD fMRI responses recorded while 8 participants each listened to 27 complete, natural, narrative stories (~6 hours). This dataset includes pre-processed and raw MRIs, as well as hand-constructed 3D cortical surfaces for each participant. To address the challenges of analyzing naturalistic data, this dataset is accompanied by a python library containing basic code for creating voxelwise encoding models. Altogether, this dataset provides a large and novel resource for understanding speech and language processing in the human brain.

Mapping cortical activations underlying covert and overt language production using high-density diffuse optical tomography

Gold standard neuroimaging modalities such as functional magnetic resonance imaging (fMRI), positron emission tomography (PET), and more recently electrocorticography (ECoG) have provided profound insights regarding the neural mechanisms underlying the processing of language, but they are limited in applications involving naturalistic language production especially in developing brains, during face-to-face dialogues, or as a brain-computer interface. High-density diffuse optical tomography (HD-DOT) provides high-fidelity mapping of human brain function with comparable spatial resolution to that of fMRI but in a silent and open scanning environment similar to real-life social scenarios. Therefore, HD-DOT has potential to be used in naturalistic settings where other neuroimaging modalities are limited. While HD-DOT has been previously validated against fMRI for mapping the neural correlates underlying language comprehension and covert (i.e., "silent") language production, HD-DOT has not yet been established for mapping the cortical responses to overt (i.e., "out loud") language production. In this study, we assessed the brain regions supporting a simple hierarchy of language tasks: silent reading of single words, covert production of verbs, and overt production of verbs in normal hearing right-handed native English speakers (n = 33). First, we found that HD-DOT brain mapping is resilient to movement associated with overt speaking. Second, we observed that HD-DOT is sensitive to key activations and deactivations in brain function underlying the perception and naturalistic production of language. Specifically, statistically significant results were observed that show recruitment of regions in occipital, temporal, motor, and prefrontal cortices across all three tasks after performing stringent cluster-extent based thresholding. Our findings lay the foundation for future HD-DOT studies of imaging naturalistic language comprehension and production during real-life social interactions and for broader applications such as presurgical language assessment and brain-machine interfaces.

The words that little by little revealed everything: Neural response to lexical-semantic content during narrative comprehension

The ease with which narratives are understood belies the complexity of the information being conveyed and the cognitive processes that support comprehension. The meanings of the words must be rapidly accessed and integrated with the reader's mental representation of the overarching, unfolding scenario. A broad, bilateral brain network is engaged by this process, but it is not clear how words that vary on specific semantic dimensions, such as ambiguity, emotion, or socialness, engage the semantic, semantic control, or social cognition systems. In the present study, data from 48 participants who listened to The Little Prince audiobook during MRI scanning were selected from the Le Petit Prince dataset. The lexical and semantic content within the narrative was quantified from the transcript words with factor scores capturing Word Length, Semantic Flexibility, Emotional Strength, and Social Impact. These scores, along with word quantity variables, were used to investigate where these predictors co-vary with activation across the brain. In contrast to studies of isolated word processing, large networks were found to co-vary with the lexical and semantic content within the narrative. An increase in semantic content engaged the ventral portion of ventrolateral ATL, consistent with its role as a semantic hub. Decreased semantic content engaged temporal pole and inferior parietal lobule, which may reflect semantic integration. The semantic control network was engaged by words with low Semantic Flexibility, perhaps due to the demand required to process infrequent, less semantically diverse language. Activation in ATL co-varied with an increase in Social Impact, which is consistent with the claim that social knowledge is housed within the neural architecture of the semantic system. These results suggest that current models of language processing may present an impoverished estimate of the neural systems that coordinate to support narrative comprehension, and, by extension, real-world language processing.

Modeling Structure-Building in the Brain With CCG Parsing and Large Language Models

To model behavioral and neural correlates of language comprehension in naturalistic environments, researchers have turned to broad-coverage tools from natural-language processing and machine learning. Where syntactic structure is explicitly modeled, prior work has relied predominantly on context-free grammars (CFGs), yet such formalisms are not sufficiently expressive for human languages. Combinatory categorial grammars (CCGs) are sufficiently expressive directly compositional models of grammar with flexible constituency that affords incremental interpretation. In this work, we evaluate whether a more expressive CCG provides a better model than a CFG for human neural signals collected with functional magnetic resonance imaging (fMRI) while participants listen to an audiobook story. We further test between variants of CCG that differ in how they handle optional adjuncts. These evaluations are carried out against a baseline that includes estimates of next-word predictability from a transformer neural network language model. Such a comparison reveals unique contributions of CCG structure-building predominantly in the left posterior temporal lobe: CCG-derived measures offer a superior fit to neural signals compared to those derived from a CFG. These effects are spatially distinct from bilateral superior temporal effects that are unique to predictability. Neural effects for structure-building are thus separable from predictability during naturalistic listening, and those effects are best characterized by a grammar whose expressive power is motivated on independent linguistic grounds.

A Texture Statistics Encoding Model Reveals Hierarchical Feature Selectivity across Human Visual Cortex

Midlevel features, such as contour and texture, provide a computational link between low- and high-level visual representations. Although the nature of midlevel representations in the brain is not fully understood, past work has suggested a texture statistics model, called the P-S model (Portilla and Simoncelli, 2000), is a candidate for predicting neural responses in areas V1-V4 as well as human behavioral data. However, it is not currently known how well this model accounts for the responses of higher visual cortex to natural scene images. To examine this, we constructed single-voxel encoding models based on P-S statistics and fit the models to fMRI data from human subjects (both sexes) from the Natural Scenes Dataset (Allen et al., 2022). We demonstrate that the texture statistics encoding model can predict the held-out responses of individual voxels in early retinotopic areas and higher-level category-selective areas. The ability of the model to reliably predict signal in higher visual cortex suggests that the representation of texture statistics features is widespread throughout the brain. Furthermore, using variance partitioning analyses, we identify which features are most uniquely predictive of brain responses and show that the contributions of higher-order texture features increase from early areas to higher areas on the ventral and lateral surfaces. We also demonstrate that patterns of sensitivity to texture statistics can be used to recover broad organizational axes within visual cortex, including dimensions that capture semantic image content. These results provide a key step forward in characterizing how midlevel feature representations emerge hierarchically across the visual system.SIGNIFICANCE STATEMENT Intermediate visual features, like texture, play an important role in cortical computations and may contribute to tasks like object and scene recognition. Here, we used a texture model proposed in past work to construct encoding models that predict the responses of neural populations in human visual cortex (measured with fMRI) to natural scene stimuli. We show that responses of neural populations at multiple levels of the visual system can be predicted by this model, and that the model is able to reveal an increase in the complexity of feature representations from early retinotopic cortex to higher areas of ventral and lateral visual cortex. These results support the idea that texture-like representations may play a broad underlying role in visual processing.

Semantic Representations during Language Comprehension Are Affected by Context

The meaning of words in natural language depends crucially on context. However, most neuroimaging studies of word meaning use isolated words and isolated sentences with little context. Because the brain may process natural language differently from how it processes simplified stimuli, there is a pressing need to determine whether prior results on word meaning generalize to natural language. fMRI was used to record human brain activity while four subjects (two female) read words in four conditions that vary in context: narratives, isolated sentences, blocks of semantically similar words, and isolated words. We then compared the signal-to-noise ratio (SNR) of evoked brain responses, and we used a voxelwise encoding modeling approach to compare the representation of semantic information across the four conditions. We find four consistent effects of varying context. First, stimuli with more context evoke brain responses with higher SNR across bilateral visual, temporal, parietal, and prefrontal cortices compared with stimuli with little context. Second, increasing context increases the representation of semantic information across bilateral temporal, parietal, and prefrontal cortices at the group level. In individual subjects, only natural language stimuli consistently evoke widespread representation of semantic information. Third, context affects voxel semantic tuning. Finally, models estimated using stimuli with little context do not generalize well to natural language. These results show that context has large effects on the quality of neuroimaging data and on the representation of meaning in the brain. Thus, neuroimaging studies that use stimuli with little context may not generalize well to the natural regime.SIGNIFICANCE STATEMENT Context is an important part of understanding the meaning of natural language, but most neuroimaging studies of meaning use isolated words and isolated sentences with little context. Here, we examined whether the results of neuroimaging studies that use out-of-context stimuli generalize to natural language. We find that increasing context improves the quality of neuro-imaging data and changes where and how semantic information is represented in the brain. These results suggest that findings from studies using out-of-context stimuli may not generalize to natural language used in daily life.

Stacked regressions and structured variance partitioning for interpretable brain maps

Relating brain activity associated with a complex stimulus to different properties of that stimulus is a powerful approach for constructing functional brain maps. However, when stimuli are naturalistic, their properties are often correlated (e.g., visual and semantic features of natural images, or different layers of a convolutional neural network that are used as features of images). Correlated properties can act as confounders for each other and complicate the interpretability of brain maps, and can impact the robustness of statistical estimators. Here, we present an approach for brain mapping based on two proposed methods: stacking different encoding models and structured variance partitioning. Our stacking algorithm combines encoding models that each use as input a feature space that describes a different stimulus attribute. The algorithm learns to predict the activity of a voxel as a linear combination of the outputs of different encoding models. We show that the resulting combined model can predict held-out brain activity better or at least as well as the individual encoding models. Further, the weights of the linear combination are readily interpretable; they show the importance of each feature space for predicting a voxel. We then build on our stacking models to introduce structured variance partitioning, a new type of variance partitioning that takes into account the known relationships between features. Our approach constrains the size of the hypothesis space and allows us to ask targeted questions about the similarity between feature spaces and brain regions even in the presence of correlations between the feature spaces. We validate our approach in simulation, showcase its brain mapping potential on fMRI data, and release a Python package. Our methods can be useful for researchers interested in aligning brain activity with different layers of a neural network, or with other types of correlated feature spaces.

Longitudinal changes in brain activation underlying reading fluency

Reading fluency-the speed and accuracy of reading connected text-is foundational to educational success. The current longitudinal study investigates the neural correlates of fluency development using a connected-text paradigm with an individualized presentation rate. Twenty-six children completed a functional MRI task in 1st/2nd grade (time 1) and again 1-2 years later (time 2). There was a longitudinal increase in activation in the ventral occipito-temporal (vOT) cortex from time 1 to time 2. This increase was also associated with improvements in reading fluency skills and modulated by individual speed demands. These findings highlight the reciprocal relationship of the vOT region with reading proficiency and its importance for supporting the developmental transition to fluent reading. These results have implications for developing effective interventions to target increased automaticity in reading.

Scaling laws for language encoding models in fMRI

Representations from transformer-based unidirectional language models are known to be effective at predicting brain responses to natural language. However, most studies comparing language models to brains have used GPT-2 or similarly sized language models. Here we tested whether larger open-source models such as those from the OPT and LLaMA families are better at predicting brain responses recorded using fMRI. Mirroring scaling results from other contexts, we found that brain prediction performance scales logarithmically with model size from 125M to 30B parameter models, with ~15% increased encoding performance as measured by correlation with a held-out test set across 3 subjects. Similar logarithmic behavior was observed when scaling the size of the fMRI training set. We also characterized scaling for acoustic encoding models that use HuBERT, WavLM, and Whisper, and we found comparable improvements with model size. A noise ceiling analysis of these large, high-performance encoding models showed that performance is nearing the theoretical maximum for brain areas such as the precuneus and higher auditory cortex. These results suggest that increasing scale in both models and data will yield incredibly effective models of language processing in the brain, enabling better scientific understanding as well as applications such as decoding.

Feature-space selection with banded ridge regression

Encoding models provide a powerful framework to identify the information represented in brain recordings. In this framework, a stimulus representation is expressed within a feature space and is used in a regularized linear regression to predict brain activity. To account for a potential complementarity of different feature spaces, a joint model is fit on multiple feature spaces simultaneously. To adapt regularization strength to each feature space, ridge regression is extended to banded ridge regression, which optimizes a different regularization hyperparameter per feature space. The present paper proposes a method to decompose over feature spaces the variance explained by a banded ridge regression model. It also describes how banded ridge regression performs a feature-space selection, effectively ignoring non-predictive and redundant feature spaces. This feature-space selection leads to better prediction accuracy and to better interpretability. Banded ridge regression is then mathematically linked to a number of other regression methods with similar feature-space selection mechanisms. Finally, several methods are proposed to address the computational challenge of fitting banded ridge regressions on large numbers of voxels and feature spaces. All implementations are released in an open-source Python package called Himalaya.

Combining computational controls with natural text reveals aspects of meaning composition

To study a core component of human intelligence-our ability to combine the meaning of words-neuroscientists have looked to linguistics. However, linguistic theories are insufficient to account for all brain responses reflecting linguistic composition. In contrast, we adopt a data-driven approach to study the composed meaning of words beyond their individual meaning, which we term 'supra-word meaning'. We construct a computational representation for supra-word meaning and study its brain basis through brain recordings from two complementary imaging modalities. Using functional magnetic resonance imaging, we reveal that hubs that are thought to process lexical meaning also maintain supra-word meaning, suggesting a common substrate for lexical and combinatorial semantics. Surprisingly, we cannot detect supra-word meaning in magnetoencephalography, which suggests that composed meaning might be maintained through a different neural mechanism than the synchronized firing of pyramidal cells. This sensitivity difference has implications for past neuroimaging results and future wearable neurotechnology.

Le Petit Prince multilingual naturalistic fMRI corpus

Neuroimaging using more ecologically valid stimuli such as audiobooks has advanced our understanding of natural language comprehension in the brain. However, prior naturalistic stimuli have typically been restricted to a single language, which limited generalizability beyond small typological domains. Here we present the Le Petit Prince fMRI Corpus (LPPC-fMRI), a multilingual resource for research in the cognitive neuroscience of speech and language during naturalistic listening (OpenNeuro: ds003643). 49 English speakers, 35 Chinese speakers and 28 French speakers listened to the same audiobook The Little Prince in their native language while multi-echo functional magnetic resonance imaging was acquired. We also provide time-aligned speech annotation and word-by-word predictors obtained using natural language processing tools. The resulting timeseries data are shown to be of high quality with good temporal signal-to-noise ratio and high inter-subject correlation. Data-driven functional analyses provide further evidence of data quality. This annotated, multilingual fMRI dataset facilitates future re-analysis that addresses cross-linguistic commonalities and differences in the neural substrate of language processing on multiple perceptual and linguistic levels.

A 10-hour within-participant magnetoencephalography narrative dataset to test models of language comprehension

Recently, cognitive neuroscientists have increasingly studied the brain responses to narratives. At the same time, we are witnessing exciting developments in natural language processing where large-scale neural network models can be used to instantiate cognitive hypotheses in narrative processing. Yet, they learn from text alone and we lack ways of incorporating biological constraints during training. To mitigate this gap, we provide a narrative comprehension magnetoencephalography (MEG) data resource that can be used to train neural network models directly on brain data. We recorded from 3 participants, 10 separate recording hour-long sessions each, while they listened to audiobooks in English. After story listening, participants answered short questions about their experience. To minimize head movement, the participants wore MEG-compatible head casts, which immobilized their head position during recording. We report a basic evoked-response analysis showing that the responses accurately localize to primary auditory areas. The responses are robust and conserved across 10 sessions for every participant. We also provide usage notes and briefly outline possible future uses of the resource.

Exploring the Representations of Individual Entities in the Brain Combining EEG and Distributional Semantics

Semantic knowledge about individual entities (i.e., the referents of proper names such as Jacinta Ardern) is fine-grained, episodic, and strongly social in nature, when compared with knowledge about generic entities (the referents of common nouns such as politician). We investigate the semantic representations of individual entities in the brain; and for the first time we approach this question using both neural data, in the form of newly-acquired EEG data, and distributional models of word meaning, employing them to isolate semantic information regarding individual entities in the brain. We ran two sets of analyses. The first set of analyses is only concerned with the evoked responses to individual entities and their categories. We find that it is possible to classify them according to both their coarse and their fine-grained category at appropriate timepoints, but that it is hard to map representational information learned from individuals to their categories. In the second set of analyses, we learn to decode from evoked responses to distributional word vectors. These results indicate that such a mapping can be learnt successfully: this counts not only as a demonstration that representations of individuals can be discriminated in EEG responses, but also as a first brain-based validation of distributional semantic models as representations of individual entities. Finally, in-depth analyses of the decoder performance provide additional evidence that the referents of proper names and categories have little in common when it comes to their representation in the brain.

A CNN-transformer hybrid approach for decoding visual neural activity into text

BACKGROUND AND OBJECTIVE

Most studies used neural activities evoked by linguistic stimuli such as phrases or sentences to decode the language structure. However, compared to linguistic stimuli, it is more common for the human brain to perceive the outside world through non-linguistic stimuli such as natural images, so only relying on linguistic stimuli cannot fully understand the information perceived by the human brain. To address this, an end-to-end mapping model between visual neural activities evoked by non-linguistic stimuli and visual contents is demanded.

METHODS

Inspired by the success of the Transformer network in neural machine translation and the convolutional neural network (CNN) in computer vision, here a CNN-Transformer hybrid language decoding model is constructed in an end-to-end fashion to decode functional magnetic resonance imaging (fMRI) signals evoked by natural images into descriptive texts about the visual stimuli. Specifically, this model first encodes a semantic sequence extracted by a two-layer 1D CNN from the multi-time visual neural activity into a multi-level abstract representation, then decodes this representation, step by step, into an English sentence.

RESULTS

Experimental results show that the decoded texts are semantically consistent with the corresponding ground truth annotations. Additionally, by varying the encoding and decoding layers and modifying the original positional encoding of the Transformer, we found that a specific architecture of the Transformer is required in this work.

CONCLUSIONS

The study results indicate that the proposed model can decode the visual neural activities evoked by natural images into descriptive text about the visual stimuli in the form of sentences. Hence, it may be considered as a potential computer-aided tool for neuroscientists to understand the neural mechanism of visual information processing in the human brain in the future.

Single-Trial MEG Data Can Be Denoised Through Cross-Subject Predictive Modeling

A pervasive challenge in brain imaging is the presence of noise that hinders investigation of underlying neural processes, with Magnetoencephalography (MEG) in particular having very low Signal-to-Noise Ratio (SNR). The established strategy to increase MEG's SNR involves averaging multiple repetitions of data corresponding to the same stimulus. However, repetition of stimulus can be undesirable, because underlying neural activity has been shown to change across trials, and repeating stimuli limits the breadth of the stimulus space experienced by subjects. In particular, the rising popularity of naturalistic studies with a single viewing of a movie or story necessitates the discovery of new approaches to increase SNR. We introduce a simple framework to reduce noise in single-trial MEG data by leveraging correlations in neural responses across subjects as they experience the same stimulus. We demonstrate its use in a naturalistic reading comprehension task with 8 subjects, with MEG data collected while they read the same story a single time. We find that our procedure results in data with reduced noise and allows for better discovery of neural phenomena. As proof-of-concept, we show that the N400m's correlation with word surprisal, an established finding in literature, is far more clearly observed in the denoised data than the original data. The denoised data also shows higher decoding and encoding accuracy than the original data, indicating that the neural signals associated with reading are either preserved or enhanced after the denoising procedure.

Generalizable EEG Encoding Models with Naturalistic Audiovisual Stimuli

In natural conversations, listeners must attend to what others are saying while ignoring extraneous background sounds. Recent studies have used encoding models to predict electroencephalography (EEG) responses to speech in noise-free listening situations, sometimes referred to as "speech tracking." Researchers have analyzed how speech tracking changes with different types of background noise. It is unclear, however, whether neural responses from acoustically rich, naturalistic environments with and without background noise can be generalized to more controlled stimuli. If encoding models for acoustically rich, naturalistic stimuli are generalizable to other tasks, this could aid in data collection from populations of individuals who may not tolerate listening to more controlled and less engaging stimuli for long periods of time. We recorded noninvasive scalp EEG while 17 human participants (8 male/9 female) listened to speech without noise and audiovisual speech stimuli containing overlapping speakers and background sounds. We fit multivariate temporal receptive field encoding models to predict EEG responses to pitch, the acoustic envelope, phonological features, and visual cues in both stimulus conditions. Our results suggested that neural responses to naturalistic stimuli were generalizable to more controlled datasets. EEG responses to speech in isolation were predicted accurately using phonological features alone, while responses to speech in a rich acoustic background were more accurate when including both phonological and acoustic features. Our findings suggest that naturalistic audiovisual stimuli can be used to measure receptive fields that are comparable and generalizable to more controlled audio-only stimuli.SIGNIFICANCE STATEMENT Understanding spoken language in natural environments requires listeners to parse acoustic and linguistic information in the presence of other distracting stimuli. However, most studies of auditory processing rely on highly controlled stimuli with no background noise, or with background noise inserted at specific times. Here, we compare models where EEG data are predicted based on a combination of acoustic, phonetic, and visual features in highly disparate stimuli-sentences from a speech corpus and speech embedded within movie trailers. We show that modeling neural responses to highly noisy, audiovisual movies can uncover tuning for acoustic and phonetic information that generalizes to simpler stimuli typically used in sensory neuroscience experiments.

Cognitive and Neural State Dynamics of Narrative Comprehension

Narrative comprehension involves a constant interplay of the accumulation of incoming events and their integration into a coherent structure. This study characterizes cognitive states during narrative comprehension and the network-level reconfiguration occurring dynamically in the functional brain. We presented movie clips of temporally scrambled sequences to human participants (male and female), eliciting fluctuations in the subjective feeling of comprehension. Comprehension occurred when processing events that were highly causally related to the previous events, suggesting that comprehension entails the integration of narratives into a causally coherent structure. The functional neuroimaging results demonstrated that the integrated and efficient brain state emerged during the moments of narrative integration with the increased level of activation and across-modular connections in the default mode network. Underlying brain states were synchronized across individuals when comprehending novel narratives, with increased occurrences of the default mode network state, integrated with sensory processing network, during narrative integration. A model based on time-resolved functional brain connectivity predicted changing cognitive states related to comprehension that are general across narratives. Together, these results support adaptive reconfiguration and interaction of the functional brain networks on causal integration of the narratives.SIGNIFICANCE STATEMENT The human brain can integrate temporally disconnected pieces of information into coherent narratives. However, the underlying cognitive and neural mechanisms of how the brain builds a narrative representation remain largely unknown. We showed that comprehension occurs as the causally related events are integrated to form a coherent situational model. Using fMRI, we revealed that the large-scale brain states and interaction between brain regions dynamically reconfigure as comprehension evolves, with the default mode network playing a central role during moments of narrative integration. Overall, the study demonstrates that narrative comprehension occurs through a dynamic process of information accumulation and causal integration, supported by the time-varying reconfiguration and brain network interaction.

The "Narratives" fMRI dataset for evaluating models of naturalistic language comprehension

The "Narratives" collection aggregates a variety of functional MRI datasets collected while human subjects listened to naturalistic spoken stories. The current release includes 345 subjects, 891 functional scans, and 27 diverse stories of varying duration totaling ~4.6 hours of unique stimuli (~43,000 words). This data collection is well-suited for naturalistic neuroimaging analysis, and is intended to serve as a benchmark for models of language and narrative comprehension. We provide standardized MRI data accompanied by rich metadata, preprocessed versions of the data ready for immediate use, and the spoken story stimuli with time-stamped phoneme- and word-level transcripts. All code and data are publicly available with full provenance in keeping with current best practices in transparent and reproducible neuroimaging.

A neural decoding algorithm that generates language from visual activity evoked by natural images

Transforming neural activities into language is revolutionary for human-computer interaction as well as functional restoration of aphasia. Present rapid development of artificial intelligence makes it feasible to decode the neural signals of human visual activities. In this paper, a novel Progressive Transfer Language Decoding Model (PT-LDM) is proposed to decode visual fMRI signals into phrases or sentences when natural images are being watched. The PT-LDM consists of an image-encoder, a fMRI encoder and a language-decoder. The results showed that phrases and sentences were successfully generated from visual activities. Similarity analysis showed that three often-used evaluation indexes BLEU, ROUGE and CIDEr reached 0.182, 0.197 and 0.680 averagely between the generated texts and the corresponding annotated texts in the testing set respectively, significantly higher than the baseline. Moreover, we found that higher visual areas usually had better performance than lower visual areas and the contribution curve of visual response patterns in language decoding varied at successively different time points. Our findings demonstrate that the neural representations elicited in visual cortices when scenes are being viewed have already contained semantic information that can be utilized to generate human language. Our study shows potential application of language-based brain-machine interfaces in the future, especially for assisting aphasics in communicating more efficiently with fMRI signals.

Stimulus-independent neural coding of event semantics: Evidence from cross-sentence fMRI decoding

Multivariate neuroimaging studies indicate that the brain represents word and object concepts in a format that readily generalises across stimuli. Here we investigated whether this was true for neural representations of simple events described using sentences. Participants viewed sentences describing four events in different ways. Multivariate classifiers were trained to discriminate the four events using a subset of sentences, allowing us to test generalisation to novel sentences. We found that neural patterns in a left-lateralised network of frontal, temporal and parietal regions discriminated events in a way that generalised successfully over changes in the syntactic and lexical properties of the sentences used to describe them. In contrast, decoding in visual areas was sentence-specific and failed to generalise to novel sentences. In the reverse analysis, we tested for decoding of syntactic and lexical structure, independent of the event being described. Regions displaying this coding were limited and largely fell outside the canonical semantic network. Our results indicate that a distributed neural network represents the meaning of event sentences in a way that is robust to changes in their structure and form. They suggest that the semantic system disregards the surface properties of stimuli in order to represent their underlying conceptual significance.

Using EEG to decode semantics during an artificial language learning task

BACKGROUND

As we learn a new nonnative language (L2), we begin to build a new map of concepts onto orthographic representations. Eventually, L2 can conjure as rich a semantic representation as our native language (L1). However, the neural processes for mapping a new orthographic representation to a familiar meaning are not well understood or characterized.

METHODS

Using electroencephalography and an artificial language that maps symbols to English words, we show that it is possible to use machine learning models to detect a newly formed semantic mapping as it is acquired.

RESULTS

Through a trial-by-trial analysis, we show that we can detect when a new semantic mapping is formed. Our results show that, like word meaning representations evoked by a L1, the localization of the newly formed neural representations is highly distributed, but the representation may emerge more slowly after the onset of the symbol. Furthermore, our mapping of word meanings to symbols removes the confound of the semantics to the visual characteristics of the stimulus, a confound that has been difficult to disentangle previously.

CONCLUSION

We have shown that the L1 semantic representation conjured by a newly acquired L2 word can be detected using decoding techniques, and we give the first characterization of the emergence of that mapping. Our work opens up new possibilities for the study of semantic representations during L2 learning.

Incremental Language Comprehension Difficulty Predicts Activity in the Language Network but Not the Multiple Demand Network

What role do domain-general executive functions play in human language comprehension? To address this question, we examine the relationship between behavioral measures of comprehension and neural activity in the domain-general "multiple demand" (MD) network, which has been linked to constructs like attention, working memory, inhibitory control, and selection, and implicated in diverse goal-directed behaviors. Specifically, functional magnetic resonance imaging data collected during naturalistic story listening are compared with theory-neutral measures of online comprehension difficulty and incremental processing load (reading times and eye-fixation durations). Critically, to ensure that variance in these measures is driven by features of the linguistic stimulus rather than reflecting participant- or trial-level variability, the neuroimaging and behavioral datasets were collected in nonoverlapping samples. We find no behavioral-neural link in functionally localized MD regions; instead, this link is found in the domain-specific, fronto-temporal "core language network," in both left-hemispheric areas and their right hemispheric homotopic areas. These results argue against strong involvement of domain-general executive circuits in language comprehension.

Incremental Language Comprehension Difficulty Predicts Activity in the Language Network but Not the Multiple Demand Network

What role do domain-general executive functions play in human language comprehension? To address this question, we examine the relationship between behavioral measures of comprehension and neural activity in the domain-general "multiple demand" (MD) network, which has been linked to constructs like attention, working memory, inhibitory control, and selection, and implicated in diverse goal-directed behaviors. Specifically, functional magnetic resonance imaging data collected during naturalistic story listening are compared with theory-neutral measures of online comprehension difficulty and incremental processing load (reading times and eye-fixation durations). Critically, to ensure that variance in these measures is driven by features of the linguistic stimulus rather than reflecting participant- or trial-level variability, the neuroimaging and behavioral datasets were collected in nonoverlapping samples. We find no behavioral-neural link in functionally localized MD regions; instead, this link is found in the domain-specific, fronto-temporal "core language network," in both left-hemispheric areas and their right hemispheric homotopic areas. These results argue against strong involvement of domain-general executive circuits in language comprehension.

Behavioral correlates of cortical semantic representations modeled by word vectors

The quantitative modeling of semantic representations in the brain plays a key role in understanding the neural basis of semantic processing. Previous studies have demonstrated that word vectors, which were originally developed for use in the field of natural language processing, provide a powerful tool for such quantitative modeling. However, whether semantic representations in the brain revealed by the word vector-based models actually capture our perception of semantic information remains unclear, as there has been no study explicitly examining the behavioral correlates of the modeled brain semantic representations. To address this issue, we compared the semantic structure of nouns and adjectives in the brain estimated from word vector-based brain models with that evaluated from human behavior. The brain models were constructed using voxelwise modeling to predict the functional magnetic resonance imaging (fMRI) response to natural movies from semantic contents in each movie scene through a word vector space. The semantic dissimilarity of brain word representations was then evaluated using the brain models. Meanwhile, data on human behavior reflecting the perception of semantic dissimilarity between words were collected in psychological experiments. We found a significant correlation between brain model- and behavior-derived semantic dissimilarities of words. This finding suggests that semantic representations in the brain modeled via word vectors appropriately capture our perception of word meanings.

Deep Artificial Neural Networks Reveal a Distributed Cortical Network Encoding Propositional Sentence-Level Meaning

Understanding how and where in the brain sentence-level meaning is constructed from words presents a major scientific challenge. Recent advances have begun to explain brain activation elicited by sentences using vector models of word meaning derived from patterns of word co-occurrence in text corpora. These studies have helped map out semantic representation across a distributed brain network spanning temporal, parietal, and frontal cortex. However, it remains unclear whether activation patterns within regions reflect unified representations of sentence-level meaning, as opposed to superpositions of context-independent component words. This is because models have typically represented sentences as "bags-of-words" that neglect sentence-level structure. To address this issue, we interrogated fMRI activation elicited as 240 sentences were read by 14 participants (9 female, 5 male), using sentences encoded by a recurrent deep artificial neural-network trained on a sentence inference task (InferSent). Recurrent connections and nonlinear filters enable InferSent to transform sequences of word vectors into unified "propositional" sentence representations suitable for evaluating intersentence entailment relations. Using voxelwise encoding modeling, we demonstrate that InferSent predicts elements of fMRI activation that cannot be predicted by bag-of-words models and sentence models using grammatical rules to assemble word vectors. This effect occurs throughout a distributed network, which suggests that propositional sentence-level meaning is represented within and across multiple cortical regions rather than at any single site. In follow-up analyses, we place results in the context of other deep network approaches (ELMo and BERT) and estimate the degree of unpredicted neural signal using an "experiential" semantic model and cross-participant encoding.SIGNIFICANCE STATEMENT A modern-day scientific challenge is to understand how the human brain transforms word sequences into representations of sentence meaning. A recent approach, emerging from advances in functional neuroimaging, big data, and machine learning, is to computationally model meaning, and use models to predict brain activity. Such models have helped map a cortical semantic information-processing network. However, how unified sentence-level information, as opposed to word-level units, is represented throughout this network remains unclear. This is because models have typically represented sentences as unordered "bags-of-words." Using a deep artificial neural network that recurrently and nonlinearly combines word representations into unified propositional sentence representations, we provide evidence that sentence-level information is encoded throughout a cortical network, rather than in a single region.

The Language Network Is Recruited but Not Required for Nonverbal Event Semantics

The ability to combine individual concepts of objects, properties, and actions into complex representations of the world is often associated with language. Yet combinatorial event-level representations can also be constructed from nonverbal input, such as visual scenes. Here, we test whether the language network in the human brain is involved in and necessary for semantic processing of events presented nonverbally. In Experiment 1, we scanned participants with fMRI while they performed a semantic plausibility judgment task versus a difficult perceptual control task on sentences and line drawings that describe/depict simple agent-patient interactions. We found that the language network responded robustly during the semantic task performed on both sentences and pictures (although its response to sentences was stronger). Thus, language regions in healthy adults are engaged during a semantic task performed on pictorial depictions of events. But is this engagement necessary? In Experiment 2, we tested two individuals with global aphasia, who have sustained massive damage to perisylvian language areas and display severe language difficulties, against a group of age-matched control participants. Individuals with aphasia were severely impaired on the task of matching sentences to pictures. However, they performed close to controls in assessing the plausibility of pictorial depictions of agent-patient interactions. Overall, our results indicate that the left frontotemporal language network is recruited but not necessary for semantic processing of nonverbally presented events.

Design of complex neuroscience experiments using mixed-integer linear programming

Over the past few decades, neuroscience experiments have become increasingly complex and naturalistic. Experimental design has in turn become more challenging, as experiments must conform to an ever-increasing diversity of design constraints. In this article, we demonstrate how this design process can be greatly assisted using an optimization tool known as mixed-integer linear programming (MILP). MILP provides a rich framework for incorporating many types of real-world design constraints into a neuroscience experiment. We introduce the mathematical foundations of MILP, compare MILP to other experimental design techniques, and provide four case studies of how MILP can be used to solve complex experimental design challenges.

Neural representation of words within phrases: Temporal evolution of color-adjectives and object-nouns during simple composition

In language, stored semantic representations of lexical items combine into an infinitude of complex expressions. While the neuroscience of composition has begun to mature, we do not yet understand how the stored representations evolve and morph during composition. New decoding techniques allow us to crack open this very hard question: we can train a model to recognize a representation in one context or time-point and assess its accuracy in another. We combined the decoding approach with magnetoencephalography recorded during a picture naming task to investigate the temporal evolution of noun and adjective representations during speech planning. We tracked semantic representations as they combined into simple two-word phrases, using single words and two-word lists as non-combinatory controls. We found that nouns were generally more decodable than adjectives, suggesting that noun representations were stronger and/or more consistent across trials than those of adjectives. When training and testing across contexts and times, the representations of isolated nouns were recoverable when those nouns were embedded in phrases, but not so if they were embedded in lists. Adjective representations did not show a similar consistency across isolated and phrasal contexts. Noun representations in phrases also sustained over time in a way that was not observed for any other pairing of word class and context. These findings offer a new window into the temporal evolution and context sensitivity of word representations during composition, revealing a clear asymmetry between adjectives and nouns. The impact of phrasal contexts on the decodability of nouns may be due to the nouns’ status as head of phrase—an intriguing hypothesis for future research.