Biology subjects

Sheth, S.

Publications and source records attributed to Sheth, S..

At least 19 recordsLinked to original sources

bioRxiv2026

A number simplex in the human medial temporal lobe

Humans handle numbers nimbly, suggesting a richer neural manifold structure than the prevalent mental number line model. In populations of medial temporal lobe (MTL) neurons in humans performing two simple tasks (dot counting and arithmetic), we find robust neural coding of numerosity that results in high dimensional, simplex-shaped manifolds. This shape affords more flexibility than a linear manifold due to its high shattering dimensionality and expressibility. Dot arrays and Arabic numerals evoked distinct simplicial population codes, yet they were linked by a linearly transferable latent structure within the same task. We find similar simplicial geometry of number representations in large language models (LLMs). Moreover, subjects internally computed arithmetic results were decodable during the calculation period, with decoding accuracy correlating with individual mathematical capacity. Finally, linear transformations of simplicial operand representations modeled the brains conversion of operands into decodable results, suggesting that the brains arithmetic procedures have some resemblance to the attention architecture of LLMs. Together, these findings establish a high dimensional representational foundation for numerical cognition in the brain.

neuroscience↗

bioRxiv2026

Polysemanticity in human hippocampal neurons

To comprehend language, the brain must navigate a high-dimensional semantic landscape while seamlessly contextualizing meaning. Inspired by recent advances in the mechanistic interpretability of large language models (LLMs), we hypothesized that the brain utilizes polysemanticity, a coding strategy wherein individual neurons represent multiple semantically unrelated features through high-dimensional superposition (Elhage et al., 2022; Olah et al., 2020). We recorded single-unit activity from the human hippocampus during podcast listening. We found that hippocampal neurons exhibit dense semantic codes characterized by multiple tuning peaks with an overdispersed, isotropic geometry. This geometry satisfies the theoretical requirements for interference minimization in superimposed codes. Furthermore, semantic responses are strongly modulated by lexical and speaker-identity context; nonetheless, the underlying population geometry remains stable. This coding strategy permits rapid contextualization without requiring specialized, context-specific neurons. Indeed, we show clear pattern separation of similar terms, along with pattern completion for held-out words. Together, these results demonstrate that the human brain leverages superposition to solve a universal computational problem: maximizing semantic capacity within a constrained representational space.

neuroscience↗

bioRxiv2026

Neural signatures of impaired semantic contextualization in Autism Spectrum Disorder

Social and communicative deficits are defining characteristics of autism spectrum disorder (ASD). Some theories suggest that these challenges, among other autistic traits, may arise from differences in predictive coding, or how the brain uses context to predict and interpret incoming information. This idea has the potential to link symptoms of autism to specific neurocomputational processes, and is especially promising for communication, whose impairment is a hallmark of ASD. Here we leveraged the ability of large language models (LLMs) to quantify semantic contextualization to analyze a unique dataset of responses from hippocampal neurons obtained during language listening in three mild-to-severe autistic individuals with comorbid epilepsy. Key elements of semantic coding were preserved in all three individuals with ASD: single-neuron response dynamics, representation of word-word semantic relationships, and patterns of context-dependent shifts in meaning. However, relative to controls, ASD resulted in reduced neural signatures of contextualization: (1) neuronal responses were aligned with earlier, less contextual layers of GPT-2, (2) ASD patients had lower effective dimensionality of the neural subspace predicting semantics, (3) neural representations of word meaning were less influenced by preceding context, and (4) neural signatures of lexical surprisal were reduced. Together, these results support theories of autism that emphasize impairments in contextualization, and highlight the power of LLMs as a tool for quantifying the computational basis of neurodevelopmental disorders.

neuroscience↗

bioRxiv2026

Semantic axes in the brain support analogical representations

ABSTRACTIn language models of word meaning, directions in the embedding space often correspond to semantic features that can be reused across different words. For example, a single direction corresponding to gender may differentiate word pairs like "boy/girl", "uncle/aunt" and "king/queen". Here we show that the same principle governs semantically driven neural responses in the human brain. We recorded populations of single neurons during podcast listening and identified word sets with consistent meaning differences. Across fifteen sets, including gender, plural, and negation, we observed consistent vectorial directions, resulting in parallelogram structures within the neural manifold. Deviation from parallelism in large language models (LLMs) predicted corresponding deviations in brain-derived parallelism. Among pronouns, vectors corresponding to case, number and person exhibited parallelogram structures individually and, collectively, obeyed the principle of commutativity, resulting in a prismatic structure. Finally, different semantic variables were preferentially associated with discrete groups of neurons, consistent with energy-efficiency theories. Together, these results establish a geometric foundation for the neural encoding of word meaning.

neuroscience↗

bioRxiv2025

Shared neural geometries for bilingual semantic representations

The human brain has the remarkable ability to comprehend and express similar concepts in multiple languages. To understand how it does so, we examined responses of hippocampal neurons during passive listening, directed speaking, and spontaneous conversation, in both English and Spanish, in a small group of balanced bilinguals. We find a small number of putative cross-language neurons, whose responses to equivalent words (e.g., "tierra" and "earth") are correlated. However, neurons semantic tunings differed substantially by language, suggesting language-specific neural implementations. Instead, the crucial driver of translation was a preserved geometric organization of neural responses between the two languages, one that did not depend on neuron level functional overlap. Indeed, that geometry was implemented by a common set of neurons along distinct readout axes; this difference in readout may help prevent cross-language interference. Together, these results suggest that hippocampus encodes a language-independent internal model for meaning.

neuroscience↗

bioRxiv2025

A semantotopic map in human hippocampus

The hippocampus plays a central role in encoding abstract conceptual and semantic information. However, little is known about the topography of that encoding. We leveraged the rare opportunity to examine neural responses densely along a small portion of the human hippocampus, specifically, the mediolateral axis of the anterior body. We collected responses to passive language listening using Neuropixels probes in three anesthetized patients during clinically indicated neurosurgical procedures. We computed semantic tuning functions for each recording site by regressing threshold crossing events and single unit responses against semantic embeddings from GPT-2, Word2Vec, and SBERT. We find that tuning functions of more distant recording sites are more dissimilar, supporting the hypothesis semantotopic organization. Multiple semantic features showed systematic changes along that axis, including animacy, concreteness, and familiarity; notably, effects were individual-specific. Surprisingly, we also found a small but significant increase in semantic similarity as a function of distance between recording sites, on a shorter spatial scale, suggesting a modest periodic organization. Together, these results demonstrate the presence of a multiscale functional organization of semantics in the hippocampus.

neuroscience↗

bioRxiv2025

The neural basis of emotional generalization in empathy

The essence of empathy is generalization of emotion across persons. Here, we leverage recent theoretical advances in the neuroscience of generalization to help us understand empathy. We measured brain activity in human neurosurgical patients performing two tasks, one focused on identifying their own emotional response and one identifying emotional responses in others. We quantified the representational geometry of local field potential (LFP) high-gamma activity in four regions: the medial temporal lobe, anterior cingulate cortex, orbitofrontal cortex, and insula. We found encoding of both self- and other-emotions in all four regions, but codes for emotion and person are disentangled (that is, factorized) in the insula, but not the other regions. This factorized representation allows for cross-person generalization of emotion in a way that tangled (non-factorized) representations do not. Together, these results support the hypothesis that the insula uniquely contributes to social mirroring processes by which we understand emotions across individuals.

neuroscience↗

bioRxiv2025

Mirror manifolds: partially overlapping neural subspaces for speaking and listening

We utilize internal representations of meaning for two purposes: to understand the words we hear and to generate our own speech. This dual requirement necessitates abstract, modality-agnostic representations. Building on work identifying it as a hub for relational mapping, we hypothesized that the hippocampus supports abstract, cross-person representations, and uses shared semantic geometries to do so. We tested this hypothesis by examining hippocampal activity in a remarkable single-neuron dataset derived from conversational speech. Neurons robustly encoded meanings of both spoken and heard words, and used common geometric embeddings for both, leading to abstract meaning performance. Speaker identity was aligned with meaning via partial subspace alignment, which affords speaker-meaning binding by partitioning meaning by speaker while maintaining cross-speaker generalization. Degrees of subspace rotation varied on a single word level and depended systematically on semantic category. Together, these findings indicate how geometric principles allow for abstract cross-personal meanings while preserving binding to speaker identity.

neuroscience↗

bioRxiv2025

Insular Traveling Waves Link Distributed Neural Dynamics to Human Memory Performance

The insula is a critical brain region that plays a foundational role in adaptive human behaviors, with diverse subregions performing distinct functional roles. However, explaining how these insular subregions interact to support behaviors is elusive. Using direct recordings from humans performing a spatial episodic memory task, we show that traveling waves within the insula modulate neuronal interactions across insula subregions, by propagating in distinct spatial patterns during specific phases of memory. In addition to traveling plane waves, insula waves also propagated in complex, heterogenous spatial patterns across task conditions. Insular traveling waves correlated with memory success, highlighting the critical role of insular traveling waves in orchestrating memory performance. Our study suggests that insular traveling waves are a key mechanism for modulating interactions and neural coding across regions to support memory processing and potentially a biomarker for investigating dysfunctions in neurological disorders.

neuroscience↗

bioRxiv2025

Cooperativity between Ras pathway mutations in colonic tumorigenesis

The Kirsten rat sarcoma (KRAS) gene is the most frequently mutated oncogene in colorectal cancer (CRC). We previously characterized two activating alleles of KRAS, A59T and A59E, that show impaired BRAF dimerization. These alleles of KRAS are enriched in CRC tumors with genetic alterations in genes that regulate MAPK signaling (e.g. EGFR, NF1). Using a new conditional mouse model of K-Ras (LSL-K-Ras A59E) we show that, despite its inability to universally activate RAF kinases, K-Ras A59E disturbs colon epithelium and decreases mouse survival in a tumor model. By combining LSL-K-Ras A59E mice with a conditional knockout of Nf1, we demonstrate cooperation between these alleles at multiple levels. Consistent with cooperation and clinical observations, we show that K-Ras A59E neither promotes EGF independence of organoid growth, nor confers intrinsic resistance to EGFR inhibition. Thus, our data provide a deeper understanding of the role of RAF isoforms in mutant KRAS driven CRC and argue for the use of anti-EGFR therapies against some Ala59 mutant alleles of KRAS. Statement of significanceOur studies on oncogenic K-RAS mutants demonstrate that the biochemical properties of a mutant oncoprotein are reflected in the somatic genetics of cancer, in particular in the cooperating mutations that occur. These mutant-specific genetic interactions influence therapeutic responses. As such, KRAS mutation may not be a univariate predictor of response to EGFR inhibition.

cancer biology↗

bioRxiv2025

Human neuronal firing is modulated by the frequency of local field potential oscillations

Neural oscillations play a critical role in shaping neuronal firing patterns. While phase-locked neuronal firing ("phase tuning") has been extensively studied in animal models and human invasive recordings, much less is known about whether neurons show preferential firing at specific oscillatory frequencies, termed frequency tuning. Here, we employ human intracranial recordings across several brain regions including hippocampus, entorhinal cortex, anterior and posterior cingulate cortex, and orbitofrontal cortex to test the hypothesis that neurons exhibit frequency-specific firing. We analyzed 357 single units recorded simultaneously with local field potentials in 19 neurosurgical patients during awake resting. We estimated the instantaneous frequency of the LFP using adaptive spectral decomposition and assessed frequency tuning of each neuron while controlling for changes in firing rate unrelated to frequency changes. We found 27% neurons exhibited increased or decreased firing within specific frequencies, most commonly within the low-theta range (<10 Hz). Neurons exhibiting frequency tuning were distinct from those displaying phase tuning, and both types of tuning were observed across multiple brain regions with no anatomical preference. Together, our results demonstrate that the instantaneous frequency of neural oscillations modulates neuronal firing which may serve as an additional mechanism for information processing in the human brain, opening new avenues for frequency-targeted neural stimulation.

neuroscience↗

bioRxiv2025

Human brain network control of creativity

The ability to think creatively is fundamental to human cognition, shaping how we interact with the world and navigate complex problems. Creativity depends on the interaction of multiple large-scale brain networks, but how these networks represent creative thought and distinguish it from other cognitive states remains unclear. Here, we use invasive intracranial recordings in the human brain to explore the network-level representations underlying creative thinking. Our results demonstrate that cortical networks in the human brain differentially encode distinct cognitive states: we found unique brain states underlying creative thinking vs. arithmetic calculations, particularly in the default mode network. Further, in the dorsal attention network, we uncovered nonlinear, high-dimensional representations of moment-by-moment fluctuations in creative performance. These representations define a neural creativity axis shared between network-level and single-neuron computations. Our findings reveal widespread neural representations of cognitive state and suggest distinct roles of specific cortical networks in controlling creativity, with the default mode network gating access to creative states and the dorsal attention network regulating the quality of creative output.

neuroscience↗

bioRxiv2025

Complementary roles for hippocampus and anterior cingulate in composing continuous choice

Naturalistic, goal directed behavior often requires continuous actions directed at dynamically changing goals. In this context, the closest analogue to choice is a strategic reweighting of multiple goal-specific control policies in response to shifting environmental pressures. To understand the algorithmic and neural bases of choice in continuous contexts, we examined behavior and brain activity in humans performing a continuous prey-pursuit task. Using a newly developed control-theoretic decomposition of behavior, we find pursuit strategies are well described by a meta-controller dictating a mixture of lower-level controllers, each linked to specific pursuit goals. We find that hippocampal neurons encode the policy blending variable in a value-invariant manner and monitor policy switches after they occur. ACC neurons encode policy switches in a value-dependent manner, with value related modulation detectable several hundred ms before the switch, alongside a ramping increase in mean firing rate toward the switch. Meanwhile, OFC activity is consistent with an encoding of the current value structure of the task, rather than policy switching. Together these results are consistent with a tripartite functional division in which hippocampus serves as a controller over behavior, ACC serves as a meta-controller, and OFC provides a value context signal. Overall, our results shed light onto the complex processes associated with choice during naturalistic continuous interactive behavior.

neuroscience↗

bioRxiv2025

Independent Continuous Tracking of Multiple Agents in the Human Hippocampus

The pursuit of fleeing prey is a core element of many species behavioral repertoires. It poses the difficult problem of continuous tracking of multiple agents, including both self and others. To understand how this tracking is implemented neurally, we examined responses of hippocampal neurons while humans performed a joystick-controlled continuous prey-pursuit task involving two simultaneously fleeing prey (and, in some cases, a predator) in a virtual open field. We found neural maps encoding the positions of all the agents. All maps were multiplexed in single neurons and were disambiguated by the use of the population coding principle of semi-orthogonal subspaces, which can facilitate cross-agent generalization. Some neurons, more common in the posterior hippocampus, had narrow tuning functions reminiscent of place cells, lower firing rates, and high information per spike; others, which were found in both anterior and posterior hippocampus, had broad tuning functions, higher firing rates, and less information per spike. Semi-orthogonalization was selectively associated with the broadly tuned neurons. These results suggest an answer to the problem of navigational individuation, that is, how mapping codes can distinguish different agents, and establish the neuronavigational foundations of pursuit.

neuroscience↗

bioRxiv2025

Parietal cortex is recruited by frontal and cingulate areas to support action monitoring and updating during stopping

Recent evidence indicates that the intraparietal sulcus (IPS) may play a causal role in action stopping, potentially representing a novel neuromodulation target for inhibitory control dysfunctions. Here, we leverage intracranial recordings in human subjects to establish the timing and directionality of information flow between IPS and prefrontal and cingulate regions during action stopping. Prior to successful inhibition, information flows primarily from the inferior frontal gyrus (IFG), a critical inhibitory control node, to IPS. In contrast, during stopping errors the communication between IPS and IFG is lacking, and IPS is engaged by posterior cingulate cortex, an area outside of the classical inhibition network and typically associated with default mode. Anterior cingulate and orbitofrontal cortex also display performance-dependent connectivity with IPS. Our functional connectivity results provide direct electrophysiological evidence that IPS is recruited by frontal and anterior cingulate areas to support action plan monitoring/updating, and by posterior cingulate during control failures. In briefFunctional connectivity between the intraparietal sulcus (IPS) and a set of frontal and cingulate regions indicates that IPS is recruited to aid inhibitory control. Control failures are associated with increased communication with posterior cingulate. IPS could be a novel and tractable neuromodulation target for control-related neuropsychiatric disorders. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=107 SRC="FIGDIR/small/640787v1_ufig1.gif" ALT="Figure 1"> View larger version (35K): org.highwire.dtl.DTLVardef@e04699org.highwire.dtl.DTLVardef@3b1ea6org.highwire.dtl.DTLVardef@15961f4org.highwire.dtl.DTLVardef@12e144f_HPS_FORMAT_FIGEXP M_FIG C_FIG HighlightsO_LIParietal cortex displays performance-dependent activity in action stopping C_LIO_LIFunctional connectivity between IPS and IFG underlies successful stopping C_LIO_LIEarly communication from ACC and OFC to IPS is also specific to successful stopping C_LIO_LICommunication from PCC to IPS is higher during lapses in control C_LI

neuroscience↗

bioRxiv2025

A vectorial code for semantics in human hippocampus

As we listen to speech, our brains track the meanings of the words we hear. Recent successes of large language models suggest that distributed population geometry can capture rich semantic relationships between words. Motivated by this idea, we hypothesized that semantic information in the brain may likewise be expressed in distributed patterns of activity across neurons, rather than in the activity of neurons narrowly tuned to a specific word. We recorded responses of hundreds of neurons in the human hippocampus while participants listened to narrative speech. We find encoding of contextual word meaning in the simultaneous activity of neurons whose individual selectivities span multiple unrelated semantic categories. Decoding and population geometry analyses revealed distinct neural coding principles for low-versus high-frequency words, likely reflecting the greater polysemy of common words. Similar to embedding vectors in semantic language models, distance between neural population responses correlates with semantic distance; however, this effect was only observed in contextual embedding models (GPT-2 and BERT), suggesting that the semantic distance effect depends critically on contextualization. Consistent with this, we find that neural population activity supports a multidimensional semantic subspace that aligns most closely with the contextual structure captured by GPT-2. Moreover, for semantically similar words, even contextual embedders showed an inverse correlation between semantic and neural distances; we attribute this pattern to the noise-mitigating benefits of contrastive coding. Ultimately, these results provide a neurocomputational account for understanding how neural populations track word meaning.

neuroscience↗

bioRxiv2025

Cell signaling pathways discovery from multi-modal data

Deciphering cell signaling pathways is key to understanding biology, disease mechanisms, and developing new therapies. Although advances in multi-omics technologies provide richer insight into signaling, the data remain high-dimensional, heterogeneous, and difficult to interpret, and current computational tools for inferring signaling pathways are limited. To address this, we developed Incytr, a method for efficient discovery of cell signaling pathways through integration of diverse data modalities, including transcriptomics, ATAC-seq, proteomics, phosphoproteomics, and kinomics. We demonstrate its application in COVID-19, Alzheimers disease, and cancer, where it successfully recovers known pathways and generates novel, cell-type-specific hypotheses supported by multiple data types. We further show how integrating Incytr-derived pathways with biomarker and drug databases can support target and drug discovery. Finally, we show that using Incytr-derived signaling pathways as training data for simple natural language processing models can deepen our understanding of cell-cell communication and immune cell dynamics, while helping identify new therapeutic targets.

bioinformatics↗

bioRxiv2025

The Single-Cell Landscape of Peripheral and Tumor-infiltrating Immune Cells in HPV- HNSCC

Head and neck squamous cell carcinoma (HNSCC) is the sixth most common cancer worldwide. HPV-negative HNSCC, which arises in the upper airway mucosa, is particularly aggressive, with nearly half of patients succumbing to the disease within five years and limited response to immune checkpoint inhibitors compared to other cancers. There is a need to further explore the complex immune landscape in HPV-negative HNSCC to identify potential therapeutic targets. Here, we integrated two single-cell RNA sequencing datasets from 29 samples and nearly 300,000 immune cells to investigate immune cell dynamics across tumor progression and lymph node metastasis. Notable shifts toward adaptative immune cell populations were observed in the 14 distinct HNSCC-associated peripheral blood mononuclear (PBMCs) and 21 tumor-infiltrating immune cells (TICs) considering disease stages. All PBMCs and TICs revealed unique molecular signatures correlating with lymph node involvement; however, broadly, TICs increased ligand expression among effector cytokines, growth factors, and interferon-related genes. Pathway analysis comparing PBMCs and TICs further confirmed active cell signaling among Monocyte-Macrophage, Dendritic cell, Natural Killer (NK), and T cell populations. Receptor-ligand analysis revealed significant communication patterns shifts among TICs, between CD8+ T cells and NK cells, showing heightened immunosuppressive signaling that correlated with disease progression. In locally invasive HPV-negative HNSCC samples, highly multiplexed immunofluorescence assays highlighted peri-tumoral clustering of exhausted CD8+ T and NK cells, alongside their exclusion from intra-tumoral niches. These findings emphasize cytotoxic immune cells as valuable biomarkers and therapeutic targets, shedding light on the mechanisms by which the HNSCC sustainably evades immune responses.

cancer biology↗