A5067100667- Neural dynamics and brain function
- Advanced Memory and Neural Computing
- Black Holes and Theoretical Physics
- Memory and Neural Mechanisms
- Neuroscience and Neuropharmacology Research
- Neural Networks and Applications
- Particle physics theoretical and experimental studies
- Cosmology and Gravitation Theories
- Face Recognition and Perception
- Functional Brain Connectivity Studies
- Visual perception and processing mechanisms
- Quantum Chromodynamics and Particle Interactions
- Neural and Behavioral Psychology Studies
- Neural Networks and Reservoir Computing
- Noncommutative and Quantum Gravity Theories
- Neurotransmitter Receptor Influence on Behavior
- Neuroscience and Neural Engineering
- Olfactory and Sensory Function Studies
- EEG and Brain-Computer Interfaces
- Stochastic processes and financial applications
- Retinal Development and Disorders
- Photoreceptor and optogenetics research
- Advanced Text Analysis Techniques
- Spacecraft and Cryogenic Technologies
- Quantum Computing Algorithms and Architecture
University of California, San Diego
2019-2025
Columbia University
2015-2024
Salk Institute for Biological Studies
2024
Brain (Germany)
2018-2021
Center for Theoretical Physics
2018
Simons Center for Geometry and Physics
2012
State University of New York
2012
Princeton University
2006-2010
In two remarkable recent papers the planar perturbative expansion was proposed for universal function of coupling appearing in dimensions high-spin operators $\mathcal{N}=4$ super Yang-Mills theory. We study numerically integral equation derived by Beisert, Eden, and Staudacher, which resums series. a confirmation anti--de Sitter-space/conformal-field-theory (AdS/CFT) correspondence, we find smooth whose leading terms at strong match results obtained semiclassical folded string spinning...
Significance Numerous studies on primates revealed the importance of hippocampus in memory formation. The rodent literature instead focused spatial representations that are observed navigation experiments. Here, we propose a simple model reconciles main findings primate and studies. assumes is system generates compressed sensory experiences using previously acquired knowledge about statistics world. These can then be memorized more efficiently. during exploration an environment, when by...
Abstract A central assumption of neuroscience is that long-term memories are represented by the same brain areas encode sensory stimuli 1 . Neurons in inferotemporal (IT) cortex represent percept visual objects using a distributed axis code 2–4 Whether and how IT neural population represents memory remains unclear. Here we examined familiar faces encoded anterior medial face patch (AM), perirhinal (PR) temporal pole (TP). In AM PR observed encoding for rotated relative to unfamiliar at long...
Abstract Normative modeling frameworks such as Bayesian inference and reinforcement learning provide valuable insights into the fundamental principles governing adaptive behavior. While these are valued for their simplicity interpretability, reliance on few parameters often limits ability to capture realistic biological behavior, leading cycles of handcrafted adjustments that prone research subjectivity. Here, we present a novel approach leveraging recurrent neural networks discover...
The curse of dimensionality plagues models reinforcement learning and decision-making. process abstraction solves this by constructing abstract variables describing features shared different specific instances, reducing enabling generalization in novel situations. Here we characterized neural representations monkeys performing a task where hidden variable described the temporal statistics stimulus-response-outcome mappings. Abstraction was defined operationally using performance decoders...
<title>Abstract</title> How do social factors impact the brain and contribute to increased alcohol drinking? We found that rank predicts drinking, where subordinates drink more than dominants. Furthermore, isolation escalates particularly impacting who display a greater increase in drinking compared Using cellular resolution calcium imaging, we show basolateral amygdala-medial prefrontal cortex (BLA-mPFC) circuit rank-dependent manner, unlike non-specific BLA activity. The BLA-mPFC becomes...
Abstract The ability to recognize familiar visual objects is critical survival. A central assumption of neuroscience that long-term memories are represented by the same brain areas encode sensory stimuli ( 1 ). Neurons in inferotemporal (IT) cortex represent percept using a distributed axis code 2–4 Whether and how IT neural population represents memory remains unclear. Here, we examined faces encoded face patch AM perirhinal cortex. We found were distinct subspace from unfamiliar faces. was...
The observation of place cells has suggested that the hippocampus plays a special role in encoding spatial information. However, cell responses are modulated by several non-spatial variables, and reported to be rather unstable. Here we propose memory model provides novel interpretation consistent with these observations. We hypothesize is device takes advantage correlations between sensory experiences generate compressed representations episodes stored memory. A simple neural network can...
Understanding the connections between artificial and biological intelligent systems can reveal fundamental principles underlying general intelligence. While many intelligence (AI) models have a neuroscience counterpart, such are largely missing in Transformer self-attention mechanism. Here, we examine relationship attention heads human episodic memory. We focus on induction heads, which contribute to in-context learning capabilities of Transformer-based large language (LLMs). demonstrate...
Synaptic plasticity is a complex phenomenon involving multiple biochemical processes that operate on different timescales. Complexity can greatly increase memory capacity when the variables characterizing synaptic dynamics have limited precision, as shown in simple retrieval problems random patterns. Here we turn to real-world problem, face familiarity detection, and show complexity be harnessed store large number of faces recognized at later time. The recognizable grows almost linearly with...
SUMMARY In order to respond appropriately threats in the environment, brain must rapidly determine whether a stimulus is important and it positive or negative, then use that information direct behavioral responses. Neurons amygdala have long been implicated valence encoding fear responses threatening stimuli, but show transient firing response these stimuli do not match timescales of associated For decades, there has logical gap how could be mediated without an ensemble representation...
Memories are stored, retained, and recollected through complex, coupled processes operating on multiple timescales. To understand the computational principles behind these intricate networks of interactions we construct a broad class synaptic models that efficiently harnesses biological complexity to preserve numerous memories. The memory capacity scales almost linearly with number synapses, which is substantial improvement over square root scaling previous models. This was achieved by...