Sophia Sanborn
Assistant Professor (Research)of Ophthalmology and, by courtesy, of Neurology
Academic Appointments
-
Assistant Professor-Research, Ophthalmology
-
Assistant Professor-Research (By courtesy), Neurology and Neurological Sciences
-
Member, Wu Tsai Neurosciences Institute
Honors & Awards
-
Graduate Research Fellowship, National Science Foundation (2015)
-
Beinecke Scholarship, The Sperry Fund (2013)
-
Duckett Prize in Philosophy, University of Maryland, College Park (2012)
Professional Education
-
Postdoctoral Scholar, University of California, Santa Barbara, Computer Science (2022)
-
Ph.D., University of California, Berkeley, Computational Neuroscience (2021)
-
B.A., University of Maryland, College Park, Philosophy & Linguistics (2013)
All Publications
-
Dual-feature selectivity enables bidirectional coding in visual cortical neurons.
eLife
2026; 15
Abstract
Sensory neurons are traditionally viewed as feature detectors that respond with an increase in firing rate to preferred stimuli while remaining unresponsive to others. Here, we identify a dual-feature encoding strategy in macaque visual cortex, wherein many neurons in areas V1 and V4 are selectively tuned to two distinct visual features-one that enhances and one that suppresses activity-around an elevated baseline firing rate. By combining neuronal recordings with functional digital twin models-deep learning-based predictive models of biological neurons-we were able to systematically identify each neuron's preferred and non-preferred features. These feature pairs served as anchors for a continuous, low-dimensional axis in natural image similarity space, along which neuronal activity varied approximately linearly. Within a single visual area, visual features that strongly or weakly activated individual neurons also had a high probability of modulating the activity of other neurons, suggesting a shared feature selectivity across the population that structures stimulus encoding. We show that this encoding strategy is conserved across species, present in both primary and lateral visual areas of mouse cortex. Dual-feature selectivity is consistent with recent anatomical evidence for feature-specific inhibitory connectivity, complementing the feature-detector principle through circuit mechanisms in which selective excitation and inhibition may together enhance the representational capacity of the neuronal population.
View details for DOI 10.7554/eLife.109861
View details for PubMedID 43001849
-
Functional bipartite invariance in mouse primary visual cortex receptive fields.
Nature neuroscience
2026
Abstract
Sensory systems support generalization by representing features that persist under input variation; however, identifying the neuronal basis of these invariances remains difficult due to high-dimensional and nonlinear neural computations. Here we leverage the inception loop paradigm, iterating between large-scale recordings, predictive models and in silico experiments with in vivo verification, to characterize neuronal invariances in mouse primary visual cortex (V1). We synthesize varied exciting inputs (VEIs), dissimilar images that drive target neurons. These VEIs revealed a new bipartite invariance: one subfield encodes a shift-tolerant high-frequency texture and the other encodes a fixed low-frequency pattern. This division aligns with object boundaries defined by spatial frequency differences in highly activating images, suggesting a contribution to segmentation. Analysis of the MICrONS dataset revealed a hierarchy of excitatory neurons in mouse V1 layers 2/3: postsynaptic neurons exhibited greater invariance than their presynaptic inputs, while neurons with lower invariance formed more connections. Together, these results provide insights and scalable methodology for mapping neuronal invariances.
View details for DOI 10.1038/s41593-026-02213-3
View details for PubMedID 41741659
View details for PubMedCentralID 2678572
- OmniMouse: Scaling properties of multi-modal, multi-task Brain Models on 150B Neural Tokens The Fourteenth International Conference on Learning Representations 2026
-
A wireless subdural-contained brain-computer interface with 65,536 electrodes and 1,024 channels.
Nature electronics
2025; 8 (12): 1272-1288
Abstract
Electrocorticography uses non-penetrating electrodes embedded in flexible substrates to record electrical activity from the surface of the brain. To use the technology to develop minimally invasive, high-bandwidth brain-computer interfaces, it will be necessary to improve the number of recording channels and the scalability of devices, which could be achieved by merging electrodes and electronics onto a single substrate. Here we report a 50-μm-thick, mechanically flexible micro-electrocorticography brain-computer interface that integrates a 256 × 256 array of electrodes, signal processing, data telemetry and wireless powering on a single complementary metal-oxide-semiconductor substrate. The device contains 65,536 recording electrodes, from which we can simultaneously record a selectable subset of up to 1,024 channels at a given time. Our chip is wirelessly powered, and when implanted below the dura, it can communicate bidirectionally with an external relay station outside the body. We show that the device can provide chronic, reliable recordings for up to two weeks in pigs and up to two months in behaving non-human primates from the somatosensory, motor and visual cortices, decoding brain signals at high spatiotemporal resolution.
View details for DOI 10.1038/s41928-025-01509-9
View details for PubMedID 42394952
View details for PubMedCentralID PMC13327737
-
Beyond Euclid: an illustrated guide to modern machine learning with geometric, topological, and algebraic structures.
Machine learning: science and technology
2025; 6 (3): 031002
Abstract
The enduring legacy of Euclidean geometry underpins classical machine learning, which, for decades, has been primarily developed for data lying in Euclidean space. Yet, modern machine learning increasingly encounters richly structured data that is inherently non-Euclidean. This data can exhibit intricate geometric, topological and algebraic structure: from the geometry of the curvature of space-time, to topologically complex interactions between neurons in the brain, to the algebraic transformations describing symmetries of physical systems. Extracting knowledge from such non-Euclidean data necessitates a broader mathematical perspective. Echoing the 19th-century revolutions that gave rise to non-Euclidean geometry, an emerging line of research is redefining modern machine learning with non-Euclidean structures. Its goal: generalizing classical methods to unconventional data types with geometry, topology, and algebra. In this review, we provide an accessible gateway to this fast-growing field and propose a graphical taxonomy that integrates recent advances into an intuitive unified framework. We subsequently extract insights into current challenges and highlight exciting opportunities for future development in this field.
View details for DOI 10.1088/2632-2153/adf375
View details for PubMedID 40755551
View details for PubMedCentralID PMC12315666
-
Dual-feature selectivity enables bidirectional coding in visual cortical neurons.
bioRxiv : the preprint server for biology
2025
Abstract
Sensory neurons are traditionally viewed as feature detectors that respond with an increase in firing rate to preferred stimuli while remaining unresponsive to others. Here, we identify a dual-feature encoding strategy in macaque visual cortex, wherein many neurons in areas V1 and V4 are selectively tuned to two distinct visual features-one that enhances and one that suppresses activity-around an elevated baseline firing rate. By combining neuronal recordings with functional digital twin models-deep learning-based predictive models of biological neurons-we were able to systematically identify each neuron's preferred and non-preferred features. These feature pairs served as anchors for a continuous, low-dimensional axis in natural image similarity space, along which neuronal activity varied approximately linearly. Within a single visual area, visual features that strongly or weakly activated individual neurons also had a high probability of modulating the activity of other neurons, suggesting a shared feature selectivity across the population that structures stimulus encoding. We show that this encoding strategy is conserved across species, present in both primary and lateral visual areas of mouse cortex. Dual-feature selectivity is consistent with recent anatomical evidence for feature-specific inhibitory connectivity, suggesting a coding strategy in which selective excitation and inhibition increase the representational capacity of the neuronal population.
View details for DOI 10.1101/2025.07.16.665209
View details for PubMedID 40777393
-
Stable, chronic in-vivo recordings from a fully wireless subdural-contained 65,536-electrode brain-computer interface device.
bioRxiv : the preprint server for biology
2025
Abstract
Minimally invasive, high-bandwidth brain-computer-interface (BCI) devices can revolutionize human applications. With orders-of-magnitude improvements in volumetric efficiency over other BCI technologies, we developed a 50-μm-thick, mechanically flexible micro-electrocorticography (μECoG) BCI, integrating a 256×256 array of electrodes, signal processing, data telemetry, and wireless powering on a single complementary metal-oxide-semiconductor (CMOS) substrate containing 65,536 recording channels, from which we can simultaneously record a selectable subset of up to 1024 channels at a given time. Fully implanted below the dura, our chip is wirelessly powered, communicating bi-directionally with an external relay station outside the body. We demonstrated chronic, reliable recordings for up to two weeks in pigs and up to two months in behaving non-human primates from somatosensory, motor, and visual cortices, decoding brain signals at high spatiotemporal resolution.
View details for DOI 10.1101/2024.05.17.594333
View details for PubMedID 38798494
View details for PubMedCentralID PMC11118429
-
Exploring the hierarchical structure of human plans via program generation.
Cognition
2024; 255: 105990
Abstract
Human behavior is often assumed to be hierarchically structured, made up of abstract actions that can be decomposed into concrete actions. However, behavior is typically measured as a sequence of actions, which makes it difficult to infer its hierarchical structure. In this paper, we explore how people form hierarchically structured plans, using an experimental paradigm with observable hierarchical representations: participants create programs that produce sequences of actions in a language with explicit hierarchical structure. This task lets us test two well-established principles of human behavior: utility maximization (i.e. using fewer actions) and minimum description length (MDL; i.e. having a shorter program). We find that humans are sensitive to both metrics, but that both accounts fail to predict a qualitative feature of human-created programs, namely that people prefer programs with reuse over and above the predictions of MDL. We formalize this preference for reuse by extending the MDL account into a generative model over programs, modeling hierarchy choice as the induction of a grammar over actions. Our account can explain the preference for reuse and provides better predictions of human behavior, going beyond simple accounts of compressibility to highlight a principle that guides hierarchical planning.
View details for DOI 10.1016/j.cognition.2024.105990
View details for PubMedID 39616822
- Harmonics of learning: Universal fourier features emerge in invariant networks The Thirty Seventh Annual Conference on Learning Theory (COLT) 2024: 3775-3797
-
The Selective G-Bispectrum and its Inversion: Applications to G-Invariant Networks
edited by Globerson, A., Mackey, L., Belgrave, D., Fan, A., Paquet, U., Tomczak, J., Zhang, C.
NEURAL INFORMATION PROCESSING SYSTEMS (NIPS). 2024
View details for Web of Science ID 001633253000069
-
Harmonics of Learning: Universal Fourier Features Emerge in Invariant Networks
edited by Agrawal, S., Roth, A.
JMLR-JOURNAL MACHINE LEARNING RESEARCH. 2023
View details for Web of Science ID 001347137103022
- Bispectral neural networks International Conference on Learning Representations (ICLR) 2023
- A General Framework for Robust G-Invariance in G-Equivariant Networks Neural Information Processing Systems (NeurIPS) 2023
-
ICML 2023 Topological Deep Learning Challenge: Design and Results
edited by Doster, T., Emerson, T., Kvinge, H., Miolane, N., Papillon, M., Rieck, B., Sanborn, S.
JMLR-JOURNAL MACHINE LEARNING RESEARCH. 2023
View details for Web of Science ID 001220893300001
-
Quantifying Extrinsic Curvature in Neural Manifolds
IEEE COMPUTER SOC. 2023: 610-619
View details for DOI 10.1109/CVPRW59228.2023.00068
View details for Web of Science ID 001055056500064
-
Efficient Neuromorphic Signal Processing with Resonator Neurons
JOURNAL OF SIGNAL PROCESSING SYSTEMS FOR SIGNAL IMAGE AND VIDEO TECHNOLOGY
2022; 94 (10): 917-927
View details for DOI 10.1007/s11265-022-01772-5
View details for Web of Science ID 000805506900001
-
Efficient Neuromorphic Signal Processing with Loihi 2
IEEE, ELECTRON DEVICES SOC & RELIABILITY GROUP. 2021: 254-259
View details for DOI 10.1109/SiPS52927.2021.00053
View details for Web of Science ID 000783958100045
-
Learning How to Generalize.
Cognitive science
2019; 43 (8): e12777
Abstract
Generalization is a fundamental problem solved by every cognitive system in essentially every domain. Although it is known that how people generalize varies in complex ways depending on the context or domain, it is an open question how people learn the appropriate way to generalize for a new context. To understand this capability, we cast the problem of learning how to generalize as a problem of learning the appropriate hypothesis space for generalization. We propose a normative mathematical framework for learning how to generalize by learning inductive biases for which properties are relevant for generalization in a domain from the statistical structure of features and concepts observed in that domain. More formally, the framework predicts that an ideal learner should learn to generalize by either taking the weighted average of the results of generalizing according to each hypothesis space, with weights given by how well each hypothesis space fits the previously observed concepts, or by using the most likely hypothesis space. We compare the predictions of this framework to human generalization behavior with three experiments in one perceptual (rectangles) and two conceptual (animals and numbers) domains. Across all three studies we find support for the framework's predictions, including individual-level support for averaging in the third study.
View details for DOI 10.1111/cogs.12777
View details for PubMedID 31446666
-
Children's representation of abstract relations in relational/array match-to-sample tasks
COGNITIVE PSYCHOLOGY
2017; 99: 17-43
Abstract
Five experiments compared preschool children's performance to that of adults and of non-human animals on match to sample tasks involving 2-item or 16-item arrays that varied according to their composition of same or different items (Array Match-to-Sample, AMTS). They establish that, like non-human animals in most studies, 3- and 4-year-olds fail 2-item AMTS (the classic relational match to sample task introduced into the literature by Premack, 1983), and that robust success is not observed until age 6. They also establish that 3-year-olds, like non-human animal species, succeed only when they are able to encode stimuli in terms of entropy, a property of an array (namely its internal variability), rather than relations among the individuals in the array (same vs. different), whereas adults solve both 2-item and 16-item AMTS on the basis of the relations same and different. As in the case of non-human animals, the acuity of 3- and 4-year-olds' representation of entropy is insufficient to solve the 2-item same-different AMTS task. At age 4, behavior begins to contrast with that of non-human species. On 16-item AMTS, a subgroup of 4-year-olds induce a categorical rule matching all-same arrays to all-same arrays, while matching other arrays (mixed arrays of same and different items) to all-different arrays. These children tend to justify their choices using the words "same" and "different." By age 4 a number of our participants succeed at 2-item AMTS, also justifying their choices by explicit verbal appeals using words for same and different. Taken together these results suggest that the recruitment of the relational representations corresponding to the meaning of these words contributes to the better performance over the preschool years at solving array match-to-sample tasks.
View details for DOI 10.1016/j.cogpsych.2017.11.001
View details for Web of Science ID 000418985300002
View details for PubMedID 29132016
-
What Do You Mean, No? Toddlers' Comprehension of Logical "No" and "Not"
LANGUAGE LEARNING AND DEVELOPMENT
2017; 13 (4): 430-450
View details for DOI 10.1080/15475441.2017.1317253
View details for Web of Science ID 000418529800005
https://orcid.org/0000-0002-1957-7067