Opens profile photo
Follow
Click to Follow adredish
A. David Redish (RedishLab)
@adredish
Scientist studying learning, memory, and decision-making. Poet and Playwright.
Minneapolis, MN USAredishlab.umn.eduJoined September 2019

A. David Redish (RedishLab)’s Tweets

All jobs have scut work. Don't let them define the job. We should teach how to ensure that one has the time to explore and think. (Like: calendar time to rest and read and think; tell others you're busy because it's in your calendar. Read a lot. Read outside your field.)
8
Show this thread
I wonder if we spend too much time teaching students how to do the scut work (writing grants, managing budgets, &c) and not enough how to make sure there is time AROUND that scut work. We should teach the joy that exists in academia despite the scut work.
1
12
Show this thread
Science is fine. We are still seeing unprecedented expansion in scientific knowledge - it's just in other fields than semiconductors and pharmaceuticals - take a look at neuroscience, AI, astrophysics, mesoeconomics to name but a few. 5/5
19
Show this thread
Increased ability to publish smaller works (not lesser, incremental science is also important!) is a consequence of cheaper production. It has become easier to publish. Other examples: UK popular novels in the 1800s & science fiction pulps in the 1950s. Again, a good thing. 3/5
1
8
Show this thread
A question for all the GPT3 people still on twitter. Has anyone tried it with non-English training sets? I'm particularly wondering how well these models do with very differently structured languages. Also, how well does it work for translation?
1
2
New books this month: 's study of the “new science of morality”; the unpublished writings of architect Colin Rowe; and essays on speculative and science fiction. Explore these and a selection of our other new and soon-to-be-published titles here:
11
Off to #SfN2022! Our posters: W Pettine SunAM WW14 New categorization model (w/ new tasks); O Calvin MonAM VV28 Anorexia & the Websurf task; S Hoffman TuePM SS1 Maze complexity, decision strategies, mPFC disruption; M Rynes WedAM XX10 Mesoscale cortical imaging & the Barnes maze
18
Beautiful work! Seems to confirm our theoretical predictions from 2007 (pubmed.ncbi.nlm.nih.gov/17638506/) tht extinction entails change in representation of state space, driven by decreases in dopamine. Reinstatement is fast due to recall of the original state. Nice when exps match theory
Quote Tweet
1/18 Our paper on reward expectation, spatial encoding, and dopamine inputs to CA1 is finally out! Huge shout out to co-first authors twitterless postdoc Seetha Krishnan and grad student Chad Heer and our former lab tech Chery Cherian. nature.com/articles/s4146
Show this thread
13
New version of arxiv.org/abs/2003.13825 now available. A thorough discussion of Theory and Modeling in neuroscience. A good first paper to discuss with incoming nsci grad students to set the stage. Also good for advanced discussion. Thx to & our many co-authors!
1
61
And for everyone saying we should read every paper on a CV to judge quality for grants, first cull of candidates, etc, I don't have that kind of time. I'm already time-overburdened. (We all are.) CV is supposed to communicate general progress in a simple and readable format.
1
6
Show this thread