Participants viewed four colored discs, two on each side of a central fixation point, while a cue identified the two colors they should retain. The discs then became colorless outlines and rotated along an invisible circle through unpredictable angles. Because the direction and distance of motion were not known in advance, accurate performance required participants to update the locations of the two items continuously rather than reconstruct their positions only when a question appeared.

At the end, one outline was highlighted, and participants indicated its remembered color by clicking on a color wheel. Behavioral results confirmed that participants generally tracked the objects as they moved. Error relative to the probed disc increased with rotation angle, with a Pearson correlation of 0.88 and p < 0.001, while error relative to the other retained disc decreased, with r = −0.89 and p < 0.001.

Model fits indicated that this pattern was mainly associated with more frequent swaps between the two remembered items. Estimated swap rates increased with movement angle, with r = 0.92 and p < 0.001, whereas modeled memory precision did not change detectably, with r = −0.02 and p = 0.93. Even at the largest rotations, average estimated swap rates were around 10%, and the average memory error on non-swap trials was 16.8 degrees.

The CDA was derived from the difference in EEG activity between electrodes contralateral and ipsilateral to the side on which the relevant objects had originally appeared. It is an electrophysiological correlate of working-memory representations and is sensitive to the number of relevant, rather than irrelevant, visual objects. The researchers organized the trials into five post-movement configurations: both discs remained in the original hemifield; one crossed the midline; both crossed; one crossed and then returned; or both completed a full circuit and returned.

Their primary model predicted that representations moving with the objects would preserve a negative CDA when both ended in the original hemifield, eliminate the lateralized difference when they occupied opposite sides, and produce a positive CDA when both ended in the opposite hemifield. Two alternatives were tested alongside that prediction. One proposed that neural representations remained anchored to their original encoding locations, so the CDA would stay negative after any amount of mental movement, perhaps weakening as memory quality declined.

The other proposed that mental movement transformed the representations into a format no longer tracked by the CDA, so the signal would disappear after a cross-midline update and would not return. The measured CDA followed the primary model closely: it began negative, reversed when both objects crossed into the opposite hemifield, and became negative again when the objects returned. Bayesian analyses favored the hemispheric-transfer model in all five conditions, although evidence for the full-circuit condition was only anecdotal, with a Bayes factor of 2.55.

Quantitative model comparisons favored the transfer interpretation over the anchored-representation account in the three conditions in which their predictions differed. The corresponding Bayes factors were 3.57 when one object crossed, 471.7 when both crossed, and 3.84 when one crossed and returned. A Bayesian mixed-effects comparison assigned 70.9% of its model weight to the transfer account, 18.5% to anchored representations, and 10.6% to the account in which the CDA disappeared after updating.

Within the favored model, estimated CDA effects were −0.56 microvolts when both objects remained in the original hemifield, 0.60 microvolts when both crossed, and −0.33 microvolts after a full circuit; the 95% highest-density intervals were −1.03 to −0.06, 0.10 to 1.06, and −0.72 to −0.001 microvolts, respectively. Pooling the three conditions that involved a midline crossing produced further evidence that the reversal was not confined to a single configuration. A one-sided cluster-based permutation test found two periods in the working-memory experiment when the post-movement CDA was significantly above zero.

Because the pooled data contained approximately equal contributions from the three crossing conditions, this indicated that the reversal was present when both objects crossed, when one crossed and returned, and when both completed a circuit. An exploratory source-level analysis was also compatible with relocation of activity, producing an odds ratio of 14.0 for reversal in the both-crossed condition compared with the conditions in which objects ended in their original hemifield. The researchers then used a second experiment to test how much of the pattern could be attributed to spatial attention.

Participants saw the same rotating outlines but were told to ignore color and track only the positions of the two cued discs. At the probe, they reported whether the highlighted disc had been one of the tracked objects. Accuracy was 94.2% with a standard deviation of 5.1%.

A negative CDA was detectable before movement, demonstrating that attentional tracking alone generated a weaker CDA-like signal, but the post-movement pattern differed quantitatively from the working-memory result. Most notably, the condition in which both objects crossed the midline did not show the positive polarity reversal observed when participants had to retain the colors. Bayesian model weights in the attention experiment were nearly evenly divided between the transfer model, at 52.2%, and the anchored-representation model, at 47.8%, while the disappearing-CDA model received negligible weight.

A hierarchical comparison found that the reversal profile relative to conditions in which objects remained in or returned to their original hemifield averaged 1.03 microvolts in the working-memory experiment and 0.57 microvolts in the attention experiment. The between-experiment difference had a 95% highest-density interval of −0.38 to 1.21 microvolts, although posterior odds of 7.3 to 1 favored a stronger modulation in the working-memory task. The CDA in the both-crossed condition was also larger in the working-memory experiment, with posterior odds of 27.1 to 1 favoring that direction of difference.

Additional analyses suggested that alpha-band oscillations and systematic eye movements could not fully explain the task-dependent CDA effect. Lateralized alpha power shifted with object positions in both experiments, consistent with spatial attention and tracking, but condition-related modulation did not differ meaningfully between tasks: the mean difference was 0.92 microvolts squared, with a 95% highest-density interval of −2.32 to 4.37. Horizontal gaze bias pointed toward the objects’ post-movement locations, but it did not differ significantly between experiments, and gaze bias did not predict CDA amplitude in most tested conditions.

Together, these results support a distinct contribution from working memory, rather than a purely attentional or oculomotor explanation for the CDA reversal. The study has important limitations. Scalp EEG has a low signal-to-noise ratio, and the physiological meaning of the CDA remains debated; it may reflect a mixture of memory, attention, indexing, or other processes rather than stored content alone.

The attention task did not formally prevent participants from partially remembering colors or engaging spatial working memory, so some implicit memory demand could have reduced the difference between the experiments. Direct decoding of object identity from EEG could test whether the hemispheric pattern reflects the remembered content itself. The findings therefore show that lateralized EEG activity dynamically follows mentally updated locations and is stronger when active working memory is required, while remaining consistent with, rather than definitive proof of, interhemispheric transfer of remembered information.

Subject of Research: Human EEG evidence for dynamic hemispheric lateralization during spatial working-memory updates

Article Title: Contralateral delay activity during dynamic spatial updates in working memory

Article References: Chernik, C., van den Berg, R., & Lundqvist, M. (2026). Contralateral delay activity during dynamic spatial updates in working memory. iScience, 29(11), Article 117916. https://doi.org/10.1016/j.isci.2026.117916

Image Credits: AI Generated

DOI: 10.1016/j.isci.2026.117916

Keywords: working memory, contralateral delay activity, electroencephalography, hemispheric transfer, spatial cognition, mental rotation, spatial attention, neural dynamics, human neuroscience, Contralateral, delay, activity

The researchers recorded the activity of hundreds of individual hippocampal neurons in patients who had electrodes implanted for clinical reasons, typically to localize seizure activity before epilepsy surgery. While the patients listened to narrative speech, the team captured the firing patterns of these neurons with millisecond precision. This rare opportunity to eavesdrop on the living human brain at single-cell resolution allowed the investigators to ask a deceptively simple question: when a person hears a word, what do hippocampal neurons actually do with its meaning?

To answer it, the team turned to an encoding model approach, a technique in which computational models of language are used to predict neural responses. Crucially, the researchers controlled for lower-level features of the speech signal, including the phonemic content of words and their grammatical roles. Even after stripping away these confounds, semantic information remained robustly encoded in the population activity. In other words, the hippocampal neurons were not merely responding to sounds or syntactic structure; they were representing what the words meant.

One of the study’s most intriguing findings concerns the selectivity of individual neurons. Rather than acting as specialized detectors for single words or narrow categories, hippocampal neurons displayed complex, mixed selectivity, responding to multiple words drawn from multiple semantic categories. This kind of distributed, high-dimensional coding has been championed in the population doctrine of cognitive neuroscience, which holds that complex cognitive content is carried not by individual cells but by patterns of activity spread across many neurons. The hippocampal semantic code fits this doctrine well: meaning emerges from the joint pattern of firing across the population, much as a chord carries information that no single note contains.

The parallels with large language models run deeper still. In models such as Word2Vec, BERT and GPT-2, words are represented as vectors in a high-dimensional space, where the distance between vectors reflects semantic similarity. The researchers found that the distance between neural population responses in the hippocampus correlated with semantic distance computed from these embedding spaces. Words with similar meanings evoked population activity patterns that were closer together in the neural space, while unrelated words produced more distant patterns. The geometry of meaning, it seems, is not unique to machines trained on oceans of text; the human hippocampus appears to organize word meanings along comparable geometric lines.

Yet the neural code was not a simple copy of any single model. Among the embedding schemes tested, neural population activity aligned most closely with GPT-2, a contextual transformer model that represents words differently depending on the surrounding sentence. This alignment suggests that hippocampal semantic representations are heavily contextualized: the same word does not always elicit the same neural pattern, but rather a pattern shaped by the linguistic moment in which it appears. Supporting this interpretation, variation in the neural response patterns across occurrences of a word correlated with polysemy measures derived from language models, meaning that words with many context-dependent senses produced more variable neural representations than words with a single, stable meaning.

The study also uncovered a subtle and technically fascinating phenomenon the authors call contrastive coding. For pairs of semantically similar words, the relationship between semantic distance and neural distance inverted: instead of nearby meanings producing nearby neural patterns, the hippocampus pushed similar representations apart. The researchers attribute this counterintuitive pattern to the noise-mitigating benefits of contrastive coding, a principle familiar from machine learning systems that learn by pushing similar examples apart to make them more distinguishable. In the brain, where neural signals are inherently noisy, deliberately separating representations of similar concepts could prevent them from being confused with one another. This echoes earlier findings that overlapping spatial memories in the hippocampus trigger representational repulsion, suggesting the structure may apply a similar differentiation strategy to meaning as it does to space and memory.

These results reshape how neuroscientists think about the hippocampus itself. For decades, the structure has been studied primarily as the engine of episodic memory, the faculty that binds the who, what, where and when of personal experience. But a growing body of evidence, including prior work showing that hippocampal neurons reactivate conceptual representations when pronouns refer back to earlier nouns, and that the region supports the flexible use of language in conversation, points to a broader role. The new findings suggest that representing word meaning during comprehension is not a peripheral sideline for the hippocampus but a core computational function, one that may provide the semantic scaffolding upon which narrative memories are built.

The methodological rigor of the study strengthens its conclusions. The team analyzed responses across thousands of words, validated their models against shuffled controls, and demonstrated that semantic embeddings explained significantly more variance in neural activity than phonetic features alone. They also showed that the semantic-neural distance relationship held across individual patients and that removing function words and pronouns did not eliminate the effect. The data and analysis code have been made publicly available, allowing other researchers to scrutinize and extend the findings, a transparency increasingly expected in high-stakes human neuroscience.

The broader implications reach toward the frontier where neuroscience and artificial intelligence converge. If the human brain and large language models have independently arrived at similar solutions for representing meaning, then the vector-space approach that underpins modern AI may capture something fundamental about how meaning must be organized to be useful. Conversely, the brain’s contrastive tricks and contextual flexibility could inspire the next generation of models. For now, the study delivers a vivid answer to an ancient question: when you hear a word, its meaning does not live in a single neuron or a single brain region, but in a shimmering, high-dimensional pattern distributed across the hippocampal population, a pattern whose geometry looks remarkably like the embeddings of a machine that learned to speak.

Subject of Research: Semantic encoding of spoken words by hippocampal neuron populations in humans

Article Title: A population code for semantics in human hippocampus

Article References: Franch, M., Mickiewicz, E. A., Belanger, J. L., Joiner, B., Katlowitz, K. A., Zhu, H., Chavez, A. G., Chericoni, A., Paulo, D. L., Goldman, A. M., Krishnan, V., Maheshwari, A., Bartoli, E., Kemmer, S., Piantadosi, S. T., Sheth, S. A., Provenza, N. R., Hayden, B. Y., & Hennig, J. A. (2026). A population code for semantics in human hippocampus. Nature Neuroscience. https://doi.org/10.1038/s41593-026-02436-4

Image Credits: AI Generated

DOI: 10.1038/s41593-026-02436-4

Keywords: hippocampus, semantics, population coding, single-neuron recordings, large language models, GPT-2, word embeddings, language comprehension, polysemy, contrastive coding, neural encoding, human neuroscience

The intellectual foundation of the new work lies in the concept of neural manifolds. When neuroscientists record from many neurons at once, they find that the activity of the population does not explore all of the dimensions theoretically available to it. Instead, the patterns of co-activation are confined to a lower-dimensional subspace, a manifold whose axes reflect the correlated structure of the circuit. Pioneering experiments by Sadtler and colleagues in 2014, using brain-computer interfaces in monkeys, demonstrated that animals could readily learn to control a cursor when the mapping between neural activity and cursor movement stayed within the intrinsic manifold of their motor cortex. When the mapping was altered to require activity patterns outside that manifold, performance collapsed and learning proceeded far more slowly. Subsequent work by Oby and colleagues in 2019 and by Golub and colleagues in 2018 extended this principle, showing that the boundaries of the manifold impose real, measurable constraints on what a brain can learn to do.

Those experiments, however, were performed with invasive recordings in animal models. Whether the same geometric principle governs learning in the intact, behaving human brain remained an open question, and a technically daunting one. Human learning is typically studied with functional magnetic resonance imaging, which measures blood-oxygen-level-dependent signals across the cortex rather than the firing of individual neurons. Busch and colleagues tackled the challenge by extracting the latent structure of whole-brain activity from imaging data, characterizing the manifold of human brain activity at a population level and then asking how that structure changes, or fails to change, as people learn new tasks. Their central result is elegantly asymmetric: learning proceeds smoothly when the required new activity patterns remain within the existing manifold, but is sharply limited when the task demands patterns that lie beyond it.

The technical machinery behind this conclusion deserves attention, because it represents a methodological bridge between two levels of analysis that have long been difficult to reconcile. Dimensionality reduction techniques, such as principal component analysis and related latent variable methods, allow researchers to compress the activity of thousands of voxels into a small number of dominant activity patterns. The geometry of these patterns, their relative positions, orientations and curvatures in the reduced space, constitutes a fingerprint of the brain’s intrinsic organization. By tracking how task-related activity trajectories move through this space during training, the researchers could quantify precisely how far a learner had to push their neural activity away from its habitual configuration. Distance within the manifold predicted rapid improvement; distance beyond the manifold predicted stagnation. The geometry was not a passive byproduct of learning. It was the variable that best explained the pace of learning itself.

This asymmetry carries profound implications for how neuroscientists think about plasticity. The traditional view of learning emphasizes synaptic modification: connections strengthen or weaken, and behavior changes accordingly. The manifold framework does not contradict that view, but it adds a crucial layer of constraint. Synaptic plasticity may be abundant and readily available, yet if the circuit’s correlated structure funnels activity along a limited set of axes, then only certain behavioral changes are cheap to produce. Changing the manifold itself, reorganizing the fundamental correlations that define it, appears to be a far more expensive operation for the brain. In effect, the brain possesses two distinct currencies of adaptation: cheap movements within its existing neural geometry, and costly renovations of that geometry. Busch and colleagues provide the strongest human evidence yet that the exchange rate between these currencies governs what we can learn and how fast.

The findings also resonate with an older literature on muscle synergies and motor primitives. Work by Berger, Gentner, Edmunds, Pai and d’Avella in 2013 showed that the motor system composes movements from a small repertoire of coordinated activation patterns rather than controlling every muscle independently. If behavior is built from a finite set of building blocks, then learning new behaviors is easiest when the desired behavior can be assembled from blocks already in stock. The manifold results generalize this intuition from muscles to cortex: the brain’s activity repertoire constrains the behaviors it can readily acquire. Interestingly, the conceptual reach of neural geometry now extends beyond neuroscience altogether. Research in artificial intelligence, notably the work of Elhage and colleagues on features and representations inside neural networks, has revealed that artificial deep networks also organize their computations along low-dimensional manifolds, and that the geometry of these representations shapes what the networks can learn. Brains and machines, it seems, confront the same geometric trade-offs.

For the growing field of brain-computer interfaces, the implications are immediate and practical. A user learning to control a robotic limb or a communication prosthesis is, in geometric terms, being asked to steer their neural activity into configurations that map onto device commands. The new results suggest that interface designers should work with the brain’s intrinsic geometry rather than against it, aligning decoder mappings with the existing manifold to shorten training times and boost performance. Conversely, when a therapeutic goal requires the brain to adopt genuinely novel activity patterns, clinicians may need to design training protocols that gradually reshape the manifold itself, perhaps through carefully staged practice that pulls the geometry outward over time. Similar reasoning applies to neurorehabilitation after stroke, where recovery may depend on whether the damaged cortex can re-express lost skills within its surviving manifold or must construct new geometric structure from the ground up.

The study also sharpens a question that has lingered since the first manifold experiments: what determines the geometry in the first place? Candidates include the statistics of a lifetime of experience, the anatomical wiring of cortical circuits, developmental programs and genetic constraints. If intrinsic geometry is largely fixed by anatomy and history, then the fastest route to enhanced learning may be to choose training tasks that sit comfortably within a person’s existing manifold, a principle that could personalize education and skill training. If, on the other hand, geometry is malleable with the right kind of experience, then identifying the activities that expand manifolds becomes a central goal for learning science. Busch and colleagues’ human imaging paradigm provides a tool for testing these possibilities, because it allows researchers to measure manifold structure repeatedly as individuals accumulate different kinds of experience.

Writing in Nature Neuroscience, Batista emphasizes the elegance of the asymmetry at the heart of the new findings: the same geometric framework that explains effortless learning also explains its limits, unifying what previously looked like separate phenomena under a single principle. The work transforms a philosophical question, why some skills come easily and others do not, into a quantitative one, measurable in the coordinates of brain activity. As imaging methods and latent variable models continue to improve, researchers may soon be able to read out an individual’s neural geometry and predict, before the first practice session, which skills that person will master quickly and which will demand a renovation of the brain’s deepest structure. Learning, on this view, is not a climb up an abstract ladder of difficulty. It is a journey through a landscape whose contours were drawn long before the journey began, and knowing the map may be the first step toward redrawing it.

Subject of Research: How the intrinsic geometry of human brain activity constrains and guides learning

Article Title: Neural geometry guides learning

Article References: Batista, A. P. (2026). Neural geometry guides learning. Nature Neuroscience. https://doi.org/10.1038/s41593-026-02442-6

Image Credits: AI Generated

DOI: 10.1038/s41593-026-02442-6

Keywords: neural manifolds, neural geometry, motor learning, functional MRI, brain-computer interfaces, neural dimensionality, cortical plasticity, skill acquisition, latent variable models, human neuroscience, Neural, geometry