Kenji Lee
@kenjilee.bsky.social
Ph.D. student studying the in vivo identification of cell types and the neural dynamics of decision making in prefrontal cortex. Chand Lab @ BU; NINDS F31 Fellow; prev. UW, Allen Inst., and U. Puget Sound. From Hawaii 🌴
I mean, from the perspective of cell types, this is very true as well! VIP+ cells contribute very little to overall activity (small footprint=hard to detect, low firing rates) but powerfully set the gain regime of mouse visual cortex.
November 25, 2025 at 11:58 PM
I mean, from the perspective of cell types, this is very true as well! VIP+ cells contribute very little to overall activity (small footprint=hard to detect, low firing rates) but powerfully set the gain regime of mouse visual cortex.
*electrophysiologist. this supports your point and is a shortcoming of the modality. Seems like a lot of the "background activity" is volume conduction of large excitatory neurons firing in synchrony given E-E neurons have the same tuning for the most part in a patch. Roozbeh has great work on this
November 25, 2025 at 11:04 PM
*electrophysiologist. this supports your point and is a shortcoming of the modality. Seems like a lot of the "background activity" is volume conduction of large excitatory neurons firing in synchrony given E-E neurons have the same tuning for the most part in a patch. Roozbeh has great work on this
This is all to say that the dimensionality often can very per dataset based on data processing, probe type, insertion variety, etc but almost no one talks about this aspect. Task variability and other behavioral covariates are equally valid I just want to mention this as an experimentalist.
November 25, 2025 at 11:00 PM
This is all to say that the dimensionality often can very per dataset based on data processing, probe type, insertion variety, etc but almost no one talks about this aspect. Task variability and other behavioral covariates are equally valid I just want to mention this as an experimentalist.
Typically top PC's often have a lot of condition-independent signal from the perspective of dPCA. Typically, a given laminar probe will hit neurons with very similar tuning which is especially true if you spike sort poorly. Threshold crossings/multi-units typically have the exact same PSTH!
November 25, 2025 at 10:59 PM
Typically top PC's often have a lot of condition-independent signal from the perspective of dPCA. Typically, a given laminar probe will hit neurons with very similar tuning which is especially true if you spike sort poorly. Threshold crossings/multi-units typically have the exact same PSTH!
I am not an expert and there’s no “experts are arguing”. It’s just straight up wrong and it’s not even close. It’s an improvement on the “rat balls” pic for sure but is still outrageously incorrect. This is not something that was solved to which I then moved the goal posts.
November 21, 2025 at 3:30 PM
I am not an expert and there’s no “experts are arguing”. It’s just straight up wrong and it’s not even close. It’s an improvement on the “rat balls” pic for sure but is still outrageously incorrect. This is not something that was solved to which I then moved the goal posts.
Still two sulci here! Areas, position, and shape all wrong as well.
November 21, 2025 at 2:39 PM
Still two sulci here! Areas, position, and shape all wrong as well.
Putting it politely, I wouldn’t consider someone recognizing G2.5 as wrong as someone with “good anatomical knowledge”. Every systems (and probably computational) neuroscientist should know this beyond undergrad. Even this schematic is outdated. The Allen Institute’s CCFv3 is the best atlas.
November 21, 2025 at 2:38 PM
Putting it politely, I wouldn’t consider someone recognizing G2.5 as wrong as someone with “good anatomical knowledge”. Every systems (and probably computational) neuroscientist should know this beyond undergrad. Even this schematic is outdated. The Allen Institute’s CCFv3 is the best atlas.
Sorry this is still ludicrously incorrect! Mouse brains are lissencephalic. These areas should be huge patches and not gyri. Like Camille said, areas still correspond to areas from a macaque. Shape is all wrong and in the wrong position. Cerebellum should be behind rather than underneath.
November 21, 2025 at 2:11 PM
Sorry this is still ludicrously incorrect! Mouse brains are lissencephalic. These areas should be huge patches and not gyri. Like Camille said, areas still correspond to areas from a macaque. Shape is all wrong and in the wrong position. Cerebellum should be behind rather than underneath.
There’s still time!
November 19, 2025 at 6:16 AM
There’s still time!
Congrats Cole!!!! Happy you’re now in Boston! See you at the climbing gym 😉
November 18, 2025 at 6:20 AM
Congrats Cole!!!! Happy you’re now in Boston! See you at the climbing gym 😉