BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//ARNI - ECPv6.15.20//NONSGML v1.0//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
X-WR-CALNAME:ARNI
X-ORIGINAL-URL:https://arni-institute.org
X-WR-CALDESC:Events for ARNI
REFRESH-INTERVAL;VALUE=DURATION:PT1H
X-Robots-Tag:noindex
X-PUBLISHED-TTL:PT1H
BEGIN:VTIMEZONE
TZID:America/New_York
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:20240310T070000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:20241103T060000
END:STANDARD
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:20250309T070000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:20251102T060000
END:STANDARD
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:20260308T070000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:20261101T060000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20250502T113000
DTEND;TZID=America/New_York:20250502T130000
DTSTAMP:20260424T064739
CREATED:20250429T152926Z
LAST-MODIFIED:20250429T152926Z
UID:1682-1746185400-1746190800@arni-institute.org
SUMMARY:CTN: Kenneth Harris
DESCRIPTION:Title: Multidimensional structure of activity in transcriptomically identified cortical cell types \nAbstract: The cerebral cortex is comprised of hundreds of distinct cell types\, connected into a network that underpins cognition.  To characterize the geometry of population activity in these cells\, we recorded from thousands of neurons simultaneously in mouse visual cortex\, and used post-hoc in situ transcriptomics to characterize their fine subtypes. We found that cortical population activity is organized around a single dimension shared between the subspace of spontaneous activity and responses to natural image stimuli\, and that a neuron’s coupling with this dimension can be predicted from a its transcriptome. \nSpontaneous population activity differed substantially with the animal’s behavioral state.  Population activity in alert states could be classified by two primary dimensions\, the first distinguishing locomoting vs. stationary periods\, and the second correlating with running speed; neurons correlated positively and negatively with both dimensions.  In non-alert states\, spontaneous activity was organized around a single dominant dimension which spontaneously oscillated\, and was close to orthogonal to the two running-related dimensions.  The coupling of neurons with this shared dimension was almost always non-negative. \nNatural image stimuli drove population activity in a subspace that overlapped the subspace of spontaneous activity only in one dimension: the dimension that was active in spontaneous oscillations of non-alert states. Even though this dimension was spontaneously oscillatory only during non-alert states\, sensory responses along this dimension were larger during running periods. This single shared dimension accounted for a larger fraction of total stimulus-related variance in inhibitory than excitatory neurons\, and in superficial than deep excitatory populations. \nThe neurons to these three dimensions could be predicted from their transcriptomes.  This prediction was weaker amongst superficial excitatory neurons\, whose strongest transcriptomic prediction was found for coupling to the oscillatory dimension. \nWe conclude that cortical activity is organized around a single dimension shared between spontaneous oscillations and natural image responses\, and hypothesize that this dimension encodes the salience of sensory or non-sensory messages broadcast from visual cortex to other cortical regions.
URL:https://arni-institute.org/event/ctn-kenneth-harris/
LOCATION:Zuckerman Institute – L5-084\, 3227 Broadway\, New York\, NY\, United States
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20250505T083000
DTEND;TZID=America/New_York:20250506T170000
DTSTAMP:20260424T064739
CREATED:20250414T161133Z
LAST-MODIFIED:20250414T161133Z
UID:1640-1746433800-1746550800@arni-institute.org
SUMMARY:Workshop on Emerging Trends in AI
DESCRIPTION:May 5th and 6th\n8:30am to 5:30pm \nPulitzer Hall – 2950 Broadway New York\, NY 10027 \nThis two-day workshop brings together leading experts in machine learning (ML) and neuroscience to examine two emerging themes: (1) the relationship between brain resilience and algorithmic robustness\, and (2) the role of ML-driven data generation in social sciences and the possible acceleration of scientific discovery. The workshop will hold two panels: one on how insights from ML and neuroscience can potentially inform each other toward the development of more resilient and robust systems\, and the second on the ethical and practical implications of synthetic data in shaping research outcomes and policy decisions. \nThe workshop will be hosted by Columbia Engineering\, the NSF AI Institute for Artificial and Natural Intelligence (ARNI)\, and Simons Institute for the Theory of Computing. \nRegister here!
URL:https://arni-institute.org/event/workshop-on-emerging-trends-in-ai/
LOCATION:Pulitzer Hall\, 2950 Broadway\, New York\, NY\, 10027\, United States
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20250507T130000
DTEND;TZID=America/New_York:20250507T140000
DTSTAMP:20260424T064739
CREATED:20250324T152647Z
LAST-MODIFIED:20250324T152647Z
UID:1594-1746622800-1746626400@arni-institute.org
SUMMARY:ARNI Continual Learning Working Group Project
DESCRIPTION:Zoom link: https://columbiauniversity.zoom.us/j/97176853843?pwd=VLZdh6yqHBcOQhdf816lkN5ByIpIsF.1
URL:https://arni-institute.org/event/arni-continual-learning-working-group-project-5/
LOCATION:CEPSR 620\, Schapiro 530 W. 120th St
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20250508T140000
DTEND;TZID=America/New_York:20250508T150000
DTSTAMP:20260424T064739
CREATED:20250501T211001Z
LAST-MODIFIED:20250505T155052Z
UID:1686-1746712800-1746716400@arni-institute.org
SUMMARY:ARNI Emerging Researchers Talk Series #3: Matteo Alleman
DESCRIPTION:Title: Discovery of categorical concepts\n\nAbstract: We seem to like reasoning in terms of discrete\, logical categories\, even in the face of continuous variation. A generous interpretation of this phenomenon is that we create abstractions about the world which enable powerful generalization to new situations — once I know what “hot-and-sour” is\, I can instantly know if I will like any dish. But state of the art learning systems\, and as far as we can tell our own brains\, use highly distributed\, continuous representations. A way that we reconcile these two ideas — abstract symbols and continuous representations — is to imagine that there are certain “component-level” directions in the vector space which are re-used across all instances which contain the component. Like the famous “king:queen :: man:woman” square from Miklov et al (2013). This provides a clear way of encoding abstract categories in continuous vector space\, but is it possible to go the other way\, and infer the categories from the vectors?\nThe first thing that may come to mind when hearing about “inferring categories” is clustering. But this has a limitation — you cannot put together “king” and “man” into a “male” category while also clustering “king” and “queen” into a “monarch” one. It also doesn’t quite seem right to say that “king” is 0.5 “male” and 0.5 “monarch” (a soft clustering)\, or that one cluster is a subset of the other (a hierarchical clustering). Instead\, I want a way to group together items into multiple clusters based on their vector representations. When data are free to belong to multiple clusters (not in a sum-to-one probabilistic way) it becomes most natural to think of the problem as a matrix factorization\, where we are trying to express our data as the product of an otherwise unconstrained binary matrix of item-cluster assignments (call it S)\, and a real-valued matrix of cluster means (call it W)\, i.e. X = SW. That is the model I will present on.\nZoom Link: https://columbiauniversity.zoom.us/j/92200158294?pwd=nGTC4FSPz2adgaOrOC0kNfY5Nr1vYq.1&jst=3
URL:https://arni-institute.org/event/arni-emerging-researchers-talk-series-3-matteo-alleman/
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20250509T113000
DTEND;TZID=America/New_York:20250509T130000
DTSTAMP:20260424T064739
CREATED:20250506T174222Z
LAST-MODIFIED:20250506T174312Z
UID:1691-1746790200-1746795600@arni-institute.org
SUMMARY:CTN: Stephanie Palmer
DESCRIPTION:Title:\nHow behavioral and evolutionary constraints sculpt early visual processing\n \nAbstract:\nBiological systems must selectively encode partial information about the environment\, as dictated by the capacity constraints at work in all living organisms. For example\, we cannot see every feature of the light field that reaches our eyes; temporal resolution is limited by transmission noise and delays\, and spatial resolution is limited by the finite number of photoreceptors and output cells in the retina. Classical efficient coding theory describes how sensory systems can maximize information transmission given such capacity constraints\, but it treats all input features equally. Not all inputs are\, however\, of equal value to the organism. Our work quantifies whether and how the brain selectively encodes stimulus features\, specifically predictive features\, that are most useful for fast and effective movements. We have shown that efficient predictive computation starts at the earliest stages of the visual system\, in the retina. We borrow techniques from statistical physics and information theory to assess how we get terrific\, predictive vision from these imperfect (lagged and noisy) component parts. In broader terms\, we aim to build a more complete theory of efficient encoding in the brain\, and along the way have found some intriguing connections between formal notions of coarse graining in biology and physics.
URL:https://arni-institute.org/event/ctn-stephanie-palmer/
LOCATION:Zuckerman Institute – L5-084\, 3227 Broadway\, New York\, NY\, United States
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20250514T130000
DTEND;TZID=America/New_York:20250514T140000
DTSTAMP:20260424T064739
CREATED:20250324T152812Z
LAST-MODIFIED:20250324T152812Z
UID:1596-1747227600-1747231200@arni-institute.org
SUMMARY:ARNI Continual Learning Working Group Project
DESCRIPTION:Zoom link: https://columbiauniversity.zoom.us/j/97176853843?pwd=VLZdh6yqHBcOQhdf816lkN5ByIpIsF.1
URL:https://arni-institute.org/event/arni-continual-learning-working-group-project-6/
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20250516T113000
DTEND;TZID=America/New_York:20250516T130000
DTSTAMP:20260424T064739
CREATED:20250513T151616Z
LAST-MODIFIED:20250513T151616Z
UID:1718-1747395000-1747400400@arni-institute.org
SUMMARY:CTN: Gaia Tavoni
DESCRIPTION:Title: A Unified Framework for Sensory Coding in Feedback-Modulated Canonical Networks \n\nAbstract: In recent decades\, the principles of neural coding have largely been studied at the level of single neurons or unimodal sensory networks. However\, brain networks interact in complex ways\, often integrating information across sensory modalities. Notably\, we lack a theoretical framework for understanding coding in interacting networks\, where information can converge from different sources. In this talk\, I will introduce a fully analytical normative framework for neural coding in feedback-modulated canonical networks\, a ubiquitous motif in the brain. In our model\, feedback is exogenous rather than endogenous to a given modality\, mediating interactions between the senses. Our theory demonstrates that predictive coding is an emergent property of efficient codes\, unifying two primary coding schemes. It further demonstrates how the computational principles of efficient and predictive coding can be implemented at the algorithmic level by a shared neural substrate\, with different network components performing distinct and interpretable mathematical operations. Finally\, our theory explains a variety of observed unimodal and multimodal sensory effects within the same normative framework and makes new predictions about the role of feedback in optimizing multimodal codes. I will conclude by showing how optimal sensory codes can be learned in biological networks through distributed Hebbian learning. Altogether\, our theory provides a unifying view of computational\, algorithmic\, and implementational principles of sensory coding in feedback-modulated canonical networks.
URL:https://arni-institute.org/event/ctn-gaia-tavoni/
LOCATION:Zuckerman Institute – L5-084\, 3227 Broadway\, New York\, NY\, United States
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20250521T130000
DTEND;TZID=America/New_York:20250521T140000
DTSTAMP:20260424T064739
CREATED:20250324T152854Z
LAST-MODIFIED:20250324T152854Z
UID:1598-1747832400-1747836000@arni-institute.org
SUMMARY:ARNI Continual Learning Working Group Project
DESCRIPTION:Zoom link: https://columbiauniversity.zoom.us/j/97176853843?pwd=VLZdh6yqHBcOQhdf816lkN5ByIpIsF.1
URL:https://arni-institute.org/event/arni-continual-learning-working-group-project-7/
LOCATION:CEPSR 620\, Schapiro 530 W. 120th St
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20250523T113000
DTEND;TZID=America/New_York:20250523T130000
DTSTAMP:20260424T064739
CREATED:20250520T163121Z
LAST-MODIFIED:20250520T173222Z
UID:1738-1747999800-1748005200@arni-institute.org
SUMMARY:CTN: Erin Rich
DESCRIPTION:Title: Dynamics of evaluation and choice in the orbitofrontal cortex\n \nAbstract: The orbitofrontal cortex is known to be important for evaluation\, choice\, and motivated behavior\, but theories on its precise role have continually evolved. In this talk\, I will discuss new perspectives on evaluation and choice in the orbitofrontal cortex in light of work from our lab and others that emphasizes information encoded by neural populations and their dynamics. I will argue that a useful framework to understand these data comes from theories of embodied decision-making\, in which moment-by-moment processing is determined by ongoing interactions between the individual and their environment.
URL:https://arni-institute.org/event/ctn-erin-rich/
LOCATION:Zuckerman Institute – L5-084\, 3227 Broadway\, New York\, NY\, United States
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20250528T140000
DTEND;TZID=America/New_York:20250528T140000
DTSTAMP:20260424T064739
CREATED:20250507T155005Z
LAST-MODIFIED:20250522T203309Z
UID:1694-1748440800-1748440800@arni-institute.org
SUMMARY:ARNI Frontier Models for Neuroscience and Behavior Working Group (Priorly: Animal Behavior): Meeting 2
DESCRIPTION:Speakers: Matt Whiteway and Mehdi Azabou \nMeeting Description: We will share progress on a benchmark using the IBL brain-wide map dataset \nWorking group Description: Advances in neurotechnology and behavioral tracking have enabled the collection of large-scale neural and behavioral datasets\, offering new opportunities to study brain function in complex settings. However\, researchers face significant challenges in integrating and analyzing data across different brain regions\, individuals\, and behavioral contexts. Inspired by recent successes in large-scale “foundation” models in natural language and computational biology\, there is growing interest in developing both neurofoundation models—which learn from diverse neural recordings—and behavior foundation models—which capture structure in high-dimensional behavioral data. These models can be used for downstream tasks such as neural encoding and decoding\, cell-type classification\, and behavioral prediction. The goal of this working group is to explore this emerging field\, address challenges in scaling these models\, identify needs for benchmarks and tools to accelerate progress\, and determine how these models can be used to answer scientific questions in both neuroscience and behavioral research. \nLocation: ZI L4-082 \nZoom Link: https://columbiauniversity.zoom.us/j/95445157877?pwd=FawUA2r6I1bxMN3W2uyok4YtF6QdpZ.1
URL:https://arni-institute.org/event/arni-frontier-models-for-neuroscience-and-behavior-working-group-priorly-animal-behavior-meeting-2/
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20250529T150000
DTEND;TZID=America/New_York:20250529T170000
DTSTAMP:20260424T064739
CREATED:20250415T195746Z
LAST-MODIFIED:20250528T134249Z
UID:1649-1748530800-1748538000@arni-institute.org
SUMMARY:ARNI WG Multi-resource-cost optimization of neural network models: Mitya Chklovskii
DESCRIPTION:Title: Can resource optimization explain neuronal morphology and placement? \nAbstract: TBD \nZoom: https://columbiauniversity.zoom.us/j/98788275902?pwd=Lnw6VtoEMdGUg0YygbkJBF3uAKgjsO.1&jst=3
URL:https://arni-institute.org/event/arni-wg-multi-resource-cost-optimization-of-neural-network-models-mitya-chklovskii/
LOCATION:Zuckerman Institute – L5-084\, 3227 Broadway\, New York\, NY\, United States
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20250530T113000
DTEND;TZID=America/New_York:20250530T130000
DTSTAMP:20260424T064739
CREATED:20250528T130320Z
LAST-MODIFIED:20250528T130320Z
UID:1769-1748604600-1748610000@arni-institute.org
SUMMARY:CTN: Shaul Druckmann
DESCRIPTION:Title: Neural dynamics of short term memory\, from mice to human speech \nAbstract: Neural dynamics represent the hard-to-interpret substrate of circuit computations. Advances in large-scale recordings have highlighted the sheer spatiotemporal complexity of circuit dynamics within and across circuits\, portraying in detail the difficulty of interpreting such dynamics and relating it to computation. Indeed\, even in extremely simplified experimental conditions\, one observes high-dimensional temporal dynamics in the relevant circuits. This complexity can be potentially addressed by the notion that not all changes in population activity have equal meaning\, i.e.\, a small change in the evolution of activity along a particular dimension may have a bigger effect on a given computation than a large change in another. We term such conditions dimension-specific computation. \nI will first briefly discuss our work on motor preparatory activity in a delayed response task in mice were we utilized neural recordings performed simultaneously with optogenetic perturbations to probe circuit dynamics. We first revealed a remarkable robustness in the detailed evolution of certain dimensions of the population activity\, beyond what was thought to be the case experimentally and theoretically. Second\, the robust dimension in activity space carries nearly all of the decodable behavioral information whereas other non-robust dimensions contained nearly no decodable information\, as if the circuit was setup to make informative dimensions stiff\, i.e.\, resistive to perturbations\, leaving uninformative dimensions sloppy\, i.e.\, sensitive to perturbations. \nI will then present unpublished work on understanding the neural dynamics underlying preparation of speech. First\, we found that as words or brief sentences are being prepared to be spoken\, we can decode not just the first phoneme to be spoken\, but rather multiple components of the speech sequence are decodable\, i.e.\, prepared in parallel. Second\, we found that unlike some previous descriptions of sequence preparation\, components were not always separated into distinct subspaces\, but rather were found in overlapping subspaces with a structured organization of their neural geometry by element and sequence position.
URL:https://arni-institute.org/event/ctn-shaul-druckmann/
LOCATION:Zuckerman Institute – L5-084\, 3227 Broadway\, New York\, NY\, United States
END:VEVENT
END:VCALENDAR