Biology subjects

Katlowitz, K.

Publications and source records attributed to Katlowitz, K..

9 recordsLinked to original sources

bioRxiv2026

Estimation of neuronal tuning for word meaning from passively recorded naturalistic speech

The ability to derive neural-level language coding models holds great scientific and clinical potential. Current approaches are limited by the scale and ethological validity of input data; applications requiring large, rare, or naturalistic samples in particular would benefit from the ability to infer neural coding from incidental everyday speech. Here we present a novel pipeline designed to leverage spontaneous and incidental naturalistic speech. This pipeline performs transcription, segmentation, and video-assisted diarization, as well as alignment and spike detection of neural data. We apply this pipeline to a dataset derived from 21 patients (6+ days each, over 800 hours and 5 million words total). We benchmark both encoding and decoding models against extensive and rare ground-truth control datasets consisting of human-curated word-level temporal alignment and manually sorted spikes. We further validate our approach by quantifying representational drift, effect of dataset size, and differences between six brain areas. Together, these findings demonstrate that incidental natural speech is sufficiently processed in the brain to enable the estimation neural-level embeddings.

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

Attention is all you need (in the brain): semantic contextualization in human hippocampus

Word meanings in language are contextualized by surrounding words. Inspired by the self-attention mechanism in transformer-based large language models (LLMs), we hypothesized that structural composition in the brain arises from combining canonical (non-contextual) word representations with those of nearby words. We analyzed single unit activity in the human hippocampus, a region involved in semantic and contextual processing, while n=10 participants listened to podcasts. We found that hippocampal neurons encoded word position within a clause, using both ordinal and frequency-domain positional encoding. Moreover, neural responses to specific words reflected both the words own lexical semantics and a weighted sum of the embeddings of preceding words. The relative weighting of these contextualizing words correlated with LLM self-attention weights. These findings suggest that contextualization in the brain makes use of vectorial shifts that have a resemblance to attentional reweighting in LLMs, and highlight the role of the mesial temporal lobe within the broader language network.

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↗

bioRxiv2020

Distinct population code for movement kinematics and changes of ongoing movements in human subthalamic nucleus

The subthalamic nucleus (STN) is theorized to globally suppress movement through connections with downstream basal ganglia structures. Current theories are supported by increased STN activity when subjects withhold an uninitiated action plan, but a critical test of these theories requires studying STN responses when an ongoing action is replaced with an alternative. Here, we perform this test using an extended reaching task with instructions to switch movement trajectory mid-action. We show that STN activity decreases during action switches, contrary to prevalent theories. Further, beta oscillations in the local field potential in STN, which are associated with movement inhibition do not show increased power or entraining of neuronal firing during switches. We report an inhomogeneous population neural code in STN, with one sub-population encoding movement kinematics and direction and another encoding unexpected action switches. We suggest an elaborate neural code in STN that contributes to planning actions and changing the plans.

neuroscience↗