Spencer LaVere Smith

@spencerlaveresmith

Prof. @ucsantabarbara.bsky.social ‪- Runs a lab slslab.org - Works on computation, neuroscience, behavior, vision, optics, imaging, 2p / multiphoton, optical computing, machine learning / AI - Blogs at - Founded @pacificoptica.bsky.social

749 Folgt615 Follower323 Posts

YouTube Website Bluesky

Ist das dein Profil?Übernimm es — dann gehört dir spencerlaveresmith.dirty.social.Profil übernehmen

Spencer LaVere Smith@spencerlaveresmith· 24d

To be clear, we welcome people who don't have PhDs too.

Spencer LaVere Smith@spencerlaveresmith· Jul 13

Ecological Visual Processing in the Mouse - with @michaelgoard.bsky.social @crisniell.bsky.social @mbeyeler.bsky.social Visual computations that are engaged in ecological contexts: active sensing, motion processing, scene analysis, distance estimation, and spatial perception. doi.org/10.1146/annu...

Spencer LaVere Smith@spencerlaveresmith· Jul 13

In fact, noise correlations (NCs or shared variability) are more stable across stimuli than signal correlations! And Yiyi showed that recurrent connectivity can support this. (3/3)

Spencer LaVere Smith@spencerlaveresmith· Jul 13

She also showed that there are functional clusters of neurons within and across cortical areas. These form high NC (noise correlation, aka shared variability) networks. Their numbers-- not their strengths-- can account for biases in cortical areas. (2/3)

Spencer LaVere Smith@spencerlaveresmith· Jul 13

Yiyi's paper shows that correlated variability (aka "noise" correlations): 1. Are significant within and across cortical visual areas (shuffle trial control) 2. Precisely reproducible at the population level 3. Scale with signal correlations 1/3 elifesciences.org/articles/97848

Spencer LaVere Smith@spencerlaveresmith· Jul 13

I like that eLife is trying different things. One thing that is tricky is having multiple versions of a preprint and an eLife reviewed preprint. Indexers like Google Scholar don't always link to the most up-to-date versions. So it's nice when they get sync'd up! elifesciences.org/articles/97848

Visual information is broadcast among cortical areas in discrete channels

Yiyi Yu, Jeffery N. Stirman, Christopher R. Dorsett, View ORCID ProfileSpencer LaVere Smith
doi: https://doi.org/10.1101/469114
Now published in eLife doi: 10.7554/eLife.97848
AbstractFull TextInfo/HistoryMetrics Preview PDF
Abstract

Among brain areas, axonal projections carry channels of information that can be mixed to varying degrees. Here, we assess the rules for the network consisting of the primary visual cortex and higher visual areas (V1-HVA) in mice. We use large field-of-view two-photon calcium imaging to measure correlated variability (i.e., noise correlations, NCs) among thousands of neurons, forming over a million unique pairs, distributed across multiple cortical areas simultaneously. The amplitude of NCs is proportional to functional connectivity in the network, and we find that they are robust, reproducible statistical measures, and are remarkably similar across stimuli, thus providing effective constraints to network models. We used these NCs to measure the statistics of functional connectivity among tuning classes of neurons in V1 and HVAs. Using a data-driven clustering approach, we identify approximately 60 distinct tuning classes found in V1 and HVAs. We find that NCs are higher between neurons from the same tuning class, both within and across cortical areas. Thus, in the V1-HVA network, mixing of channels is avoided. Instead, distinct channels of visual information are broadcast within and across cortical areas, at both the micron and millimeter length scales. This principle for the functional organization and correlation structure at the individual neuron level across multiple cortical areas can inform and constrain computational theories of neocortical networks.

Spencer LaVere Smith@spencerlaveresmith· Jun 15

NIH BRAIN Initiative tool developers: Best wishes on your grant applications!

Spencer LaVere Smith@spencerlaveresmith· Jun 13

If neurons across the pop. are highly correlated, it could be neuropil contamination. And if it is bad enough, after neuropil subtraction the neuron signal can be lost in the shot noise. So minimize neuropil contamination (small PSFs and/or sparse labeling). doi.org/10.1038/nn.4... (3/3)

Spencer LaVere Smith@spencerlaveresmith· Jun 13

Signal detection theory plus photon counting statistics provide a fundamental estimate of the number of photons needed. Wilt et al. 2013 walks through the calculations nicely. doi.org/10.1016/j.bp... (2/3)

Spencer LaVere Smith@spencerlaveresmith· Jun 11

If you're not replacing your PMTs every 6 months, you might as well opt for SiPMs. src: www.nature.com/articles/s41... sipm: opg.optica.org/oe/fulltext.... www.nature.com/articles/s41...

Spencer LaVere Smith@spencerlaveresmith· Jun 11

2-photon optics pros: Here's a job in Alipasha's lab at the Rockefeller in NYC. careers.rockefeller.edu/jobs/2040?la... Work on new multi-photon high-speed large-scale in vivo multi-photon optical neurotechnologies!

Spencer LaVere Smith@spencerlaveresmith· Jun 11

Great paper! Large scale 2p + behav. reveals that time codes in M2 and PPC can shift between coherent and ind. modes. Plus, their shared pop. dyn. are stratified: time info in a high-D subspace, and behav info in a low-D subspace. @riichirohira.bsky.social www.nature.com/articles/s41...

Spencer LaVere Smith@spencerlaveresmith· Jun 2

grants.nih.gov/news-events/... NIH guidance on what is and isn't a foreign component. (June 1, 2026)

May Be a Foreign Component
Foreign site recruitment
Foreign specimen / data collection
Use of foreign lab facilities
Work performed by a foreign collaborator

May Not Be a Foreign Component
Foreign travel for consultation
A collaborator who moved overseas after the work was completed
Foreign vendors providing routine services
Foreign co-author providing minor editorial feedback

Spencer LaVere Smith@spencerlaveresmith· May 30

Mythbusters is my kids’ favorite show (10 yo and 6 yo). The show was canceled before they were born, but it resonates. Empiricism. Even above rigor, empiricism. Thank you to everyone who helped create this show. labrigger.com/blog/2014/01...

Spencer LaVere Smith@spencerlaveresmith· May 27

We have insurance for shipping for @pacificoptica.bsky.social and the policy covers some unusual topics, but not $10 billion telescopes. www.nasa.gov/universe/how... The logistics are fascinating and include everything from spherical aberration to pirates. www.theatlantic.com/science/arch... (2/2)

Spencer LaVere Smith@spencerlaveresmith· May 27

When the James Webb space telescope was moved from California to French Guiana, it was in an unmarked container (to avoid PIRATES!) which functioned as a mobile clean room: "no more than 100 airborne particles greater than or equal to 0.5 microns in size" www.freightwaves.com/news/james-w... (1/2)

Spencer LaVere Smith@spencerlaveresmith· May 23

Me: I see you need a new imaging system. Would you like an LLSM, ExLLSM, LLSM-PAINT, SIM, ISM, LLS-SIM, SMLM/SPT, AO-ISM, AO-LLSM, or an AO-LLS-SIM? You: Yes. Me: Good. Here are the pieces. Beautiful work by Upadhyayula, Legant, Betzig, and many other excellent people. doi.org/10.1038/s415...

Spencer LaVere Smith@spencerlaveresmith· May 23

This is Humphrey. I guess I’ll do work on the laptop some other time. #caturday

Spencer LaVere Smith@spencerlaveresmith· May 2

That’s terrifying, Instagram.

Spencer LaVere Smith@spencerlaveresmith· Apr 24

Open-source modular FPGA system for two-photon mesoscope: multiarea, multidepth neural activity recording. @riichirohira.bsky.social et al. present high throughput 2p calcium and what is maybe the largest field-of-view 2p fluorescence lifetime imaging (FLIM) data. doi.org/10.1117/1.NP...

Spencer LaVere Smith@spencerlaveresmith· Apr 23

Spencer LaVere Smith@spencerlaveresmith· Apr 18

This is beautiful data. 2p optical reporting of subthreshold voltage, validated with simultaneous patch clamp. www.biorxiv.org/content/10.6... FORCE1s by François St-Pierre's group, with a bunch of great people joining in.

Figure 2. FORCE1s reports in vivo neuronal activity with high fidelity under resonance
scanning 2-photon microscopy
(A-I) 2P-guided whole-cell patching and resonant-scan imaging in anesthetized mice.
(A) Experimental schematic.
(B) 2P images of a representative FORCE1s-Kv-expressing neuron are shown at high resolution (i) and the size and
resolution used during voltage imaging (ii). The patching pipette is outlined.
(C) Simultaneous electrophysiological recording (top, 10 kHz) and FORCE1s-Kv fluorescence imaging (bottom, 440
Hz). The fluorescence trace was low-pass filtered at 150 Hz to enhance the visualization of the spikes. Dashed
boxes: epochs shown in panels D-E. Vertical lines: detected spikes. Gray triangle (left, and in panels D-E): -70 mV.
(D-E) Zoomed-in subthreshold (D) and bursting (E) epochs. Panel D’s optical trace was low-pass filtered at 50 Hz to
enhance the visualization of subthreshold fluctuations.

Spencer LaVere Smith@spencerlaveresmith· Apr 15

Open-hardware microscopy builds & industry. Online workshop. Thursday April 16, 10:30am Central time US. Free. Industry involvement in open hardware. www.bioimagingnorthamerica.org/events/abrf-... from BioImaging North America BINA.

This event explores the connections between industry and open-hardware microscopy builds. Our speakers will share diverse perspectives, including an industry provider supporting open-hardware initiatives, an academic founder who transitioned from open builds to startups, and entrepreneurs developing commercial microscope platforms. The challenges, opportunities, and synergies that arise at the intersection of open hardware and industry will be discussed. The talks will be of interest to microscope builders, open-hardware developers, academic researchers considering the startup pathway, and community members interested in the evolving industry landscape surrounding open-hardware microscopy. The program will take place on Zoom and will feature four 15-minute presentations followed by a 30-minute interactive Q&A session. Attendees are encouraged to network with the speakers and continue the discussion within the BINA community.

Spencer LaVere Smith@spencerlaveresmith· Mar 21

“by opening the door to irrational medicine alongside evidence-based medicine, we are poisoning the minds of the public” “if we don’t put a brake on the increasing self-confidence [...] they will cease to limit their attention to self-limiting or nonspecific maladies” labrigger.com/blog/2026/03...

In image of a passage from an old and worn copy of Holmes' piece on homeopathy. It reads, "As one humble member of a profession that for more than two thousand years has devoted itself to the pursuit of the best earthly interests of mankind, always assailed and insulted from without by such as are ignorant of its infinite perplexities and labors, always striving in unequal contest with the hundred-armed giant who walks in the noonday, and sleeps not in the midnight, yet still toiling, not merely for itself and the present moment, but for the race and the future, I have lifted my voice against this lifeless delusion, rolling its shapeless bulk into the path of a noble science it is too weak to strike, or to injure."

Spencer LaVere Smith@spencerlaveresmith· Mar 20

Nice day today.

Spencer LaVere Smith@spencerlaveresmith· Mar 18

Relatedly, GCaMP7s could be more widely used-- slow decay like 6s, but with a bigger change in fluorescence. Any good transgenic mouse lines? Dana et al. 2019 www.nature.com/articles/s41... (as always, YMMV)

Spencer LaVere Smith@spencerlaveresmith· Mar 18

In this respect, GCaMP6s is still interesting. If your cells are mostly bursting, GCaMP6s' ultra long decay might get you the photons you need. I tried to plot both on the same graph (green dotted lines). So GCaMP6s is still relevant! www.nature.com/articles/nat... (3/3)

Spencer LaVere Smith@spencerlaveresmith· Mar 18

The decay time also matters. Longer decays give more photons. Fast indicators track calcium dynamics better, but if you're just trying to detect bursts, that long decay can be great. Especially if your frame rate is slow. (2/3) www.cell.com/biophysj/ful...

Spencer LaVere Smith@spencerlaveresmith· Mar 18

To get good two-photon calcium imaging signals, you need photons. So you want bright indicators, like the GCaMP8 series. www.nature.com/articles/s41... HOWEVER, that's not the whole story... (1/3)

Spencer LaVere Smith@spencerlaveresmith· Mar 11

This depends on the monitor you're using, right?