Kenneth Latimer

@latimerkw

Statistical neuroscience postdoc at UChicago working in Dave Freedman's lab.

Vrijeme pridruživanja: srpanj 2015.

Tweetovi

Blokirali ste korisnika/cu @latimerkw

Jeste li sigurni da želite vidjeti te tweetove? Time nećete deblokirati korisnika/cu @latimerkw

  1. proslijedio/la je Tweet
    16. sij

    A member made a very unsettling discovery in her transcript yesterday. If you’ve taught in the past few years, has the admin retroactively added a course with departmental code PTPT, and a title like “Graduate Student Lectureship,” to your transcript too? Take a look!

    Prikaži ovu nit
    Poništi
  2. proslijedio/la je Tweet
    3. sij

    Since my colleague (and office-next-door-neighbor) Astrid Prinz isn't on Twitter, here's an ad for a postdoc position in her group! Interested in computational modeling of neural circuits? Come join us in the hallway of computational neuroscience Biology!

    Poništi
  3. proslijedio/la je Tweet
    2. sij

    The history of vision research, in 25 papers. Check out this great review by my (twitterless) colleagues Stuart Trenholm and Arjun Krishnaswamy

    Poništi
  4. proslijedio/la je Tweet
    2. sij

    Had a ton of fun working on this new paper with the newly minted Dr. Nathan Buerkle! A wonderful wrap to 2019

    Poništi
  5. 31. pro 2019.

    My latest preprint with looking at adaptation to variance in single neurons with GLMs is up on biorxiv.

    Poništi
  6. 26. pro 2019.

    Some really fascinating results linking functional characterizations of optically recorded synaptic inputs and EM reconstructions of the synapses to challenge theories of Hebbian learning's role in shaping selectivity in Ferret V1

    Poništi
  7. proslijedio/la je Tweet
    24. pro 2019.

    SCOOP: Bloomberg’s campaign was using prison labor to make campaign calls. They said they canceled the contract yesterday, when I asked them about it. My latest for

    Poništi
  8. 19. pro 2019.

    Real-world applications of machine learning can have serious undesirable consequences that hurt people's lives. These systems shouldn't be deployed by engineers without rigorous input from social scientists and ethicists.

    Poništi
  9. 18. pro 2019.

    Excited to have my latest work with and Fred Rieke at eLife. We use a simple biophysical extension of GLMs to look at how spiking depends on the tuning of excitation and inhibition in primate RGCs.

    Poništi
  10. proslijedio/la je Tweet
    2. pro 2019.

    These were the last lectures for my BDA course. You can find all lecture videos and other course material listed at . I hope someone finds these useful even if they are far from perfect.

    Poništi
  11. 21. stu 2019.

    Glad to hear the state is responding to this report. Hopefully they'll follow through with real change.

    Prikaži ovu nit
    Poništi
  12. 20. stu 2019.

    Locking kids in isolation rooms for misbehaving in school is a monstrous practice that is somehow widespread and legal in Illinois. I hope will push to end this.

    Prikaži ovu nit
    Poništi
  13. proslijedio/la je Tweet

    Much of what’s being sold as "AI" today is snake oil. It does not and cannot work. In a talk at MIT yesterday, I described why this happening, how we can recognize flawed AI claims, and push back. Here are my annotated slides:

    Prikaži ovu nit
    Poništi
  14. proslijedio/la je Tweet
    17. stu 2019.

    Cross-validation with confidence, by Jing Lei. New cross-validation method that takes into account the uncertainty in the testing sample, yielding consistent variable selection.

    Poništi
  15. proslijedio/la je Tweet
    16. stu 2019.

    First paper from postdoc is on ! We've identified a new region in temporal cortex that reflects both attentional modulation and object selectivity. But what's even cooler is that these high-level properties depend critically on midbrain SC

    Prikaži ovu nit
    Poništi
  16. proslijedio/la je Tweet

    "In contrast to similar work in the primate, we find no match between the hierarchy of mouse visual cortical areas and the layers of CNNs trained on object categorization. Although VGG16 achieves state-of-the-art performance, it is matched by random weights."

    Prikaži ovu nit
    Poništi
  17. proslijedio/la je Tweet

    Leaked audio shines a light on the behind-the-scenes tactics beverage and plastics companies like Coca-Cola have quietly used for decades to evade responsibility for their waste.

    Poništi
  18. 5. stu 2019.

    Tensor lasso methods (with some motivation from neuroscience)

    Poništi
  19. proslijedio/la je Tweet

    No one was supposed to find out about this.

    Poništi
  20. proslijedio/la je Tweet
    28. lis 2019.

    Great review of time series forecasting and "The Importance of Being Uncertain." Explaining the past is not the same as forecasting the future. Explanatory variables are not as helpful as you'd think. And the surprisingly poor performance of machine learning methods.

    Poništi

Čini se da učitavanje traje već neko vrijeme.

Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.

    Možda bi vam se svidjelo i ovo:

    ·