Princeton Neuroscience Institute · Falkner Lab

Jorge Iravedra-García

Systems and computational neuroscientist. I build optical tools to record brain-wide circuits in freely moving mice, and models that tie network dynamics to the structure of social behavior.

Portrait of Jorge Iravedra-García
Simulated · 12 sites
12

optical fibers in a single custom implant

23

genetically identified populations recorded per animal

500+

longitudinal recording sessions

About

I am a postdoctoral research associate in the Falkner Lab at the Princeton Neuroscience Institute, where I completed my PhD in 2026.

I work on social cognition from the outside in: how the brain parses the observable features of others — what they are doing, who they are, how an encounter is going — and integrates that with internal state to arrive at a behavioral policy. Observation is a useful entry point, because a watcher has to track several individuals with opposing goals at once and work out what any of it means for itself. I approach this by recording brain-wide circuit activity in freely behaving mice and pairing it with unsupervised behavioral quantification and encoding models that link network dynamics to behavior frame by frame.

My postdoctoral work turns that recording platform into a causal one. Adding galvanometer-steered optogenetic stimulation to the same fiber bundle lets me drive one population at a time while recording from every other site, which yields a signed, directed map of effective connections across the social behavior network in an animal that is awake and behaving. The question is whether experience and observation rewrite the causal weights themselves, not only the activity those weights carry.

I started in neuroscience as a BP-ENDURE fellow at the University of Puerto Rico, Río Piedras, in Gregory Quirk's lab, recording single units in behaving rats to work out how prefrontal projections control active avoidance. That program exists to open research careers to students like me, and the community of boricua neuroscientists it connected me to is a large part of why I am still here.

Research

Selected publications. Full list on the CV and Google Scholar.

loading

Twelve recording sites

Hover or pick a region

AP
—
ML
—
Depth
—

Direct aggression experience activates the network broadly; the notes below say what each population did, and flag the exceptions. Coordinates are relative to bregma, and depth is measured from the brain surface. Red marks hypothalamic sites, violet the amygdala and BNST, tan the forebrain, brown the habenula and midbrain.

    Surface: Allen Mouse Brain Common Coordinate Framework, via the Scalable Brain Atlas

    Multifiber implant fabrication, surgical implantation, the imaging path, and isosbestic motion-artifact correction
    JoVE · 2026 · Methods

    Two-color multifiber photometry recordings of the social behavior network in mice

    Guthman, E.M.*, Iravedra-García, J.M.*, Cherepanova, P., Yokoyama, S., Lockwood, S.E., Oline, S.N. & Falkner, A.L. *Equal contribution

    We describe how to build and implant custom multifiber arrays that record population calcium activity from up to 12 sites across the mouse social behavior network in a single freely moving animal. Green- and red-shifted calcium indicators let us record two genetically defined populations at every site simultaneously, imaged with a commercial CMOS-based photometry system during social behavior. The protocol covers implant fabrication and surgery, quality control, and a preprocessing pipeline for noise correction and standardization, taking a lab from bench to analysis-ready signals from deep brain structures.

    Schematic: an observer watches a dominant and a subordinate, then updates its social structure and its inferred reward functions
    Curr Opin Neurobiol · 2026 · Review

    Watching the watchers: aggression observation as a window into the neural mechanisms and computations underlying social learning

    Tsin, D., Iravedra-García, J.M. & Falkner, A.L.

    Watching others fight changes later behavior, and recent work shows that observation recruits subcortical systems for aggressive motivation, arousal and reward, beyond classic action-mirroring circuits. We argue that aggression observation is an unusually demanding form of social learning, because the observer must infer the distinct goals of a winner and a loser at once, and propose multi-agent and inverse reinforcement learning as frameworks for formalizing how watchers update their model of the social world.

    Recording ERα+ and ERα− populations across the social behavior network and the resulting context-specific activity maps
    bioRxiv · 2025 · Preprint

    Context rescales a social action code in a hormone-sensitive network

    Guthman, E.M., Iravedra-García, J.M., Sirrs, L., Le, A. & Falkner, A.L.

    We recorded ERα+ and ERα− populations across the subcortical social behavior network while mice met different partners in different territories, and quantified behavior frame by frame without supervision. Each context produced a distinct pattern of social action, and territory linearly rescaled the partner-specific action code, amplifying behavior-associated activity at home. In males, gonadectomy disrupted home-territory behavior, which testosterone replacement rescued, and impaired territorial rescaling in a population-specific manner; these behavioral effects were absent in females.

    CV

    jmig@princeton.edu · [PDF CV link]

    Experience

    • 2026 – nowPostdoctoral research associateFalkner Lab, Princeton Neuroscience Institute
    • 2020 – 2026Graduate researcherFalkner Lab, Princeton Neuroscience Institute
    • 2017 – 2019Undergraduate researcherQuirk Lab, University of Puerto Rico School of Medicine

    Education

    • 2019 – 2026Ph.D., Neuroscience, Princeton UniversityDissertation: Brain-wide neural mechanisms underlying the observational learning of aggression. Advisor: Annegret Falkner
    • 2022M.A., Neuroscience, Princeton University
    • 2015 – 2019B.S., Molecular and Cellular BiologyUniversity of Puerto Rico, Río Piedras

    Publications

    Fellowships & awards

    • 2024Best Talk Award, PNI Annual Retreat
    • 2020NSF GRFP, Honorable Mention
    • 2019President's Fellowship, Princeton University
    • 2017 – 2019NIH BP-ENDURE Fellowship, University of Puerto Rico

    Presentations

    • 2025Cosyne, MontréalPoster
    • 2024PNI Annual RetreatTalk · Best Talk Award
    • 2022, 2023Society for NeurosciencePoster
    • 2018SACNAS National ConferenceTalk / poster

    In the media

    Radio, podcast and press appearances.

    NPR · All Things Considered

    Mice can learn to fight by watching

    Aired 9 November 2026. Full segment and transcript at NPR.

    0:00 / 3:44

    Coverage

    • Nov 2026Princeton Neuroscience Institute“Mice that watch fights come away battle-ready”
    • Nov 2026Nautilus“What happened when neuroscientists started a mouse fight club”
    • Nov 2026NPR, All Things Considered“Mice can learn to fight by watching tiny rodent cage matches, study finds”
    • Nov 2026Ars Technica“Watching mouse fight club teaches the critters how to fight”
    • Nov 2026Vice“The real reason why humans love watching other people fight, according to science”
    • Dec 2024WPRB News and CulturePodcast episode: “Kicking & Screaming”