David Aagten-Murphy

[David Aagten-Murphy @DAagtenMurphy]

I wanted to talk a little about my new paper with Paul Bays (@BaysLab) in @PLOSCompBiol in which we demonstrated that remembering the absolute (egocentric) and relative (allocentric) location of objects uses independent working memory resources [1/10]
https://t.co/xPHI2pUybq

[Tim Paris @tiimparis]

https://t.co/ll7NZbAh3I is hiring a data scientist👩‍🔬💾

Do you want to work at a place where machine learning is a core technology, not a buzzword dreamed up by marketing?

Learn more here 👇
https://t.co/DsYAkNgCkO

[Paul Bays @BaysLab]

We measured how effectively observers can reallocate working memory resources to new visual items when old ones become obsolete - people are surprisingly good at it! New paper with @ivntmc @dataforyounz @DAagtenMurphy https://t.co/EOS8mgu3LV

[David Aagten-Murphy @DAagtenMurphy]

@boeslab @willjharrison reminds me of a patient that talked about once?

[Aryeh Kontorovich @aryehazan]

@srchvrs I think the easiest way to see it is that pointwise convergence of f says nothing about convergence of f'.
Take f to be the area and f' to be the curve you're integrating.
Of course, we can have infinite length enclosing finite area (e.g., Koch) which is even "worse" than π=4

[David Aagten-Murphy @DAagtenMurphy]

@willjharrison @ReubenRideaux Cool! Colour coding the edge angles, I like it. Quite neat seeing the localized "vertical" and "horizontal" orientation parts of the photo pop out in colour. Can you share any hints about what you two are up to (for a curious bystander)?

[Anna Shpektor @AnnaShpektor]

https://t.co/z2VUKPM9LV
It is hard to find time to write every day, but here is a report about one of the days.
Every morning I rent ~8 rooms at a hotel 100m away from the station. It allows to provide comfortable sleep for ~30 people a day.

[David Aagten-Murphy @DAagtenMurphy]

What a great opportunity! Work alongside a great supervisor and teacher on a PHD project where you learn a ton of different skills (python, eye-movements, drug, trials) AND work on a topic that opens up academic, clinical, and industry paths for later?! Can definitely recommend

[David Aagten-Murphy @DAagtenMurphy]

@katohla @opodo Oh no. Worst nightmare, especially in these times. Do better @opodo

[David Aagten-Murphy @DAagtenMurphy]

@calihu Neat paper Carolin!

[David Aagten-Murphy @DAagtenMurphy]

@Twisted_Ramble just launched their Kickstarter campaign for Duru. It is a really cool "serious game" which has the player solve puzzles with mole rats - while exploring the impact of stigma in depression.
Check them out and help spread the word!
https://t.co/UKUoHLW6WV

[David Aagten-Murphy @DAagtenMurphy]

Great opportunity to work with a great team!

[Tim Paris @tiimparis]

https://t.co/ll7NZbAh3I is hiring a data scientist👩‍🔬💾

Do you want to work at a place where machine learning is a core technology, not a buzzword dreamed up by marketing?

Learn more here 👇
https://t.co/DsYAkNgCkO

[David Aagten-Murphy @DAagtenMurphy]

@jesstaubert & @susanwardle - thought you two would appreciate these!

[Ronald A. Senack Facial Pareidolia in Trees @SenackRon]

Pareidolia in trees #pareidolia #faces

[Ronald A. Senack Facial Pareidolia in Trees @SenackRon]

Pareidolia in trees #pareidolia #faces

[David Aagten-Murphy @DAagtenMurphy]

@AndyTWoods @lewislepton I would love to see this too. Give me a shout if you come across it (I will do the same)

[David Aagten-Murphy @DAagtenMurphy]

@HeedLab @BaysLab I'd be curious to see if asking subjects to sometimes recall the LM would work because now the LM would not only be a LM, but also a target. We actually tried testing a version where any stimuli could come back as a LM, but it became complicated designing adequate controls.

[David Aagten-Murphy @DAagtenMurphy]

@HeedLab @BaysLab I imagine other forms of memory / response strategies - such as "somewhere in the top left quadrant" or guessing randomly but avoiding the locations of the other explicitly remembered stimuli - might enable people to do better than chance.

[David Aagten-Murphy @DAagtenMurphy]

@HeedLab Hopefully, that address the question. Obviously, this is a bit of speculation and goes beyond the data we currently have but it should give you an idea of how I am thinking about these cues. I speak for myself here, Paul's (@BaysLab) ideas may vary! More experiments are needed!

[David Aagten-Murphy @DAagtenMurphy]

@HeedLab Presumably, if the landmark moves TOO far, the estimate reconstructed from the relative information anchored to this moved landmark would conflict with the absolute target estimate. Then (like ventriloquism with sound and vision) I expect we get competition and one might dominate