tag:blogger.com,1999:blog-70125610222692036832024-02-07T18:13:24.045-08:00Brain on Brains Anonymoushttp://www.blogger.com/profile/13266163205297209914noreply@blogger.comBlogger9125tag:blogger.com,1999:blog-7012561022269203683.post-18333623052763123092016-10-27T04:30:00.000-07:002016-10-27T07:22:13.313-07:00Where'd that memory go?<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal;">
<span style="-webkit-font-kerning: none;"><span style="font-size: large;"><b>Where'd that memory go?</b></span></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal;">
<span style="-webkit-font-kerning: none;"><span style="font-size: large;"><br /></span></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal;">
<span style="-webkit-font-kerning: none;"><span style="font-size: large;">Almost anyone, upon reading the word <i>StarWars, </i>will have one memory or another pop into their heads: ‘<<i>purhh>, I. Am. Your father.’ </i>What happened in your brain when this memory was first acquired? A recent <a href="http://biorxiv.org/content/early/2016/10/14/081018" target="_blank">preprint by Chris Baldassano</a> and colleagues put forth compelling evidence for a comprehensive theory for how the brain segments experience into events and how these events are stored in memory.</span></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal; min-height: 13px;">
<span style="font-size: large;"><span style="font-kerning: none;"></span><br /></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal;">
<span style="font-kerning: none;"><span style="font-size: large;">One of the predictions of this theory is that encoding (i.e. the storage of an event into memory) happens as discrete events. Rather than recording instant by instant your brain stores chunks of time at a time. This got me wondering, could I pick out brain activity that corresponds to these encoding events?</span></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal; min-height: 13px;">
<span style="font-size: large;"><span style="font-kerning: none;"></span><br /></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal;">
<span style="font-kerning: none;"><span style="font-size: large;">A method termed intersubject correlation (ISC) gives us a way of seeing which brain regions, between two different brains, had similar activity. ISC computes the correlation between the same pair of voxels in two different brains. Because the two brains are being submitted to the same stimulus, brain regions with larger correlations mean that activity in that region is likely being driven by the stimulus. </span></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal; min-height: 13px;">
<span style="font-size: large;"><span style="font-kerning: none;"></span><br /></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal;">
<span style="font-kerning: none;"><span style="font-size: large;">So I downloaded the publicly available <a href="http://studyforrest.org/" target="_blank">Forrest Gump dataset</a> (a dataset consisting of peoples’ brain scan while they watched the movie Forrest Gump), wrote a short<a href="https://github.com/andrebeu/Neuro-Gump/blob/master/preprocess_video_alt.sh" target="_blank"> preprocessing and analysis script</a> in AFNI and python, and borrowing from <a href="http://www.nature.com/articles/ncomms12141" target="_blank">Simoney et al. 2016</a> computed the sliding window ISC. Sliding window ISC computes the correlation between the same voxel in two brains in a small window of time, and then slides that window over and repeats. Then, naturally, I made movie visualizations of my results:</span></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal;">
<span style="font-kerning: none;"><span style="font-size: large;"><br /></span></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal;">
<span style="font-kerning: none;"><span style="font-size: large;">These are the average of the sliding ISC between one subject and three others. Note that for the purposes of this analysis I only took a small section of the brain, the hippocampal regions, well known for its involvement in memory. </span></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal;">
<br /></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal;">
<div class="separator" style="clear: both; text-align: center;">
<iframe allowfullscreen='allowfullscreen' webkitallowfullscreen='webkitallowfullscreen' mozallowfullscreen='mozallowfullscreen' width='320' height='266' src='https://www.blogger.com/video.g?token=AD6v5dzClInTW1US6c_JYXXdwZBrq8Oo7hjmPZq4bq2_Lk4dqXWIWlUknJyRNBb8S5czoesBmY3Nt53J-DE3VYQ_Qg' class='b-hbp-video b-uploaded' frameborder='0'></iframe></div>
<div class="separator" style="clear: both; text-align: center;">
<br /></div>
<iframe allowfullscreen='allowfullscreen' webkitallowfullscreen='webkitallowfullscreen' mozallowfullscreen='mozallowfullscreen' width='320' height='266' src='https://www.blogger.com/video.g?token=AD6v5dwjUfpKL8GKMUENACLmSxXjUXVMGxPHReQcQUfORGuCozVGtmL7aTnhMYocS2JaAQFo8lKSKDx7Hm3zfv9a5A' class='b-hbp-video b-uploaded' frameborder='0'></iframe><br />
<div class="separator" style="clear: both; text-align: center;">
<br /></div>
<span style="font-kerning: none;"><span style="font-size: large;">It would be hasty and ‘unsciency’ of me to draw any conclusions. But.. I am a philosophy major as well.. So I’ll say this: the little clusters of activity that transiently emerge might happen to be the neural signature of encoding events. If that turns out to be the case, I could pull out the activity in those brain regions and ask: what does the time course of an encoding event look like? Is it just an increase in activity or something more complex and nuanced?</span></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal; min-height: 13px;">
<span style="font-size: large;"><span style="font-kerning: none;"></span><br /></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal;">
<span style="font-kerning: none;"><span style="font-size: large;">To really test the hypothesis that these are encoding events, I could (should and might) take the activity timecourse in the regions the Baldassano et al. manuscript predicts triggers encoding events, take the activity in these hippocampal regions my analysis picks up and calculate some form of sliding correlation but for matrices. A sliding frobenius norm or a mantel test maybe?</span></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal; min-height: 13px;">
<span style="font-size: large;"><span style="font-kerning: none;"></span><br /></span></div>
<br />
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal;">
<span style="font-kerning: none;"><span style="font-size: large;">Anyway, gotta run, late for class. </span></span></div>
<div>
<span style="font-kerning: none;"><br /></span></div>
Anonymoushttp://www.blogger.com/profile/13266163205297209914noreply@blogger.com0tag:blogger.com,1999:blog-7012561022269203683.post-58177472469398931222016-07-23T06:48:00.001-07:002016-07-23T06:58:05.463-07:00Cognitive neuroscience: What my mom thinks I do<div class="page" title="Page 1">
<div class="section" style="background-color: rgb(100.000000%, 100.000000%, 100.000000%);">
<div class="layoutArea">
<div class="column">
<div style="text-align: center;">
<span style="font-family: "helvetica"; font-weight: 700;"><span style="font-size: large;"><br /></span></span></div>
<div style="text-align: center;">
<span style="font-family: "helvetica"; font-weight: 700;"><span style="font-size: large;">Cognitive Neuroscience:
</span></span></div>
<div style="text-align: center;">
<span style="font-family: "helvetica"; font-weight: 700;"><span style="font-size: large;">What my mom thinks I do, what my friends think I do, </span></span></div>
<div style="text-align: center;">
<span style="font-family: "helvetica"; font-weight: 700;"><span style="font-size: large;">what I thought I’d be doing, what I actually do.
</span></span></div>
<div style="text-align: center;">
<span style="font-family: "helvetica"; font-weight: 700;"><span style="font-size: large;"><br /></span></span></div>
<span style="font-family: "helvetica";"><span style="font-size: large;">I love it when I get the opportunity to say: “I am a cognitive scientist.” </span></span><br />
<span style="font-family: "helvetica";"><span style="font-size: large;">(“Co- what?” asks grandma)
</span></span><br />
<span style="font-family: "helvetica";"><span style="font-size: large;"><br /></span></span>
<span style="font-family: "helvetica";"><span style="font-size: large;">Cognitive neuroscience is the field that studies how the brain does what it does. First
and foremost the brain is a biological organ. But unlike other organs, like hearts or
kidneys, the deeds of the brain don’t have a straightforward mechanical explanation (it
doesn’t pump, it doesn’t filter). What does the brain do? How does it do what it does?
</span></span><br />
<span style="font-family: "helvetica";"><span style="font-size: large;"><br /></span></span>
<span style="font-family: "helvetica";"><span style="font-size: large;">Cognitive neuroscience is as much of a philosophical enterprise as it is a scientific one.
Whatever you do (walk, think, listen, sleep), there is a lot going on ‘behind the scenes’
that we’re typically unaware of. Part of the job is asking what is going on behind the
scenes? When that is known (to the extent possible), the goal is then to relate the
‘behind the scenes’ to the ‘scenes’ themselves.
</span></span><br />
<span style="font-family: "helvetica";"><span style="font-size: large;"><br /></span></span>
<span style="font-family: "helvetica";"><span style="font-size: large;">Being a cognitive neuroscientist sounds fun, doesn’t it? It gets even better.
</span></span><br />
<span style="font-family: "helvetica";"><span style="font-size: large;"><br /></span></span>
<span style="font-family: "helvetica";"><span style="font-size: large;">To be honest, my title is slightly misleading. I haven’t the slightest clue what my mother
and friends might think I do. As for my past self, when I decided to drop music and go
into neuroscience instead, the plan was to become the basement neuroscientist who
would eventually figure out how to get perfect memory. After all, with the wisdom of a 19
year old, I decided the time was ripe for Nietzsche’s Zarathustra. Then I could go back
to music - and no time would be lost because I would have perfect memory.</span></span><br />
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEi1dcjhPIKDYpSEyZgm1SMEa7GkDJgKRCiu2ds6h5wHFbWbgTfK51Crl35qcUosu0a6Wrtt9EVEznD8NRT4XIE_Xd3HaM-PKOWZA2M3dCHLe-XVNYMLs3kGmzi_BRuvCUquN8bQ998Jbvc/s1600/1007-5841-brain1_cover.jpg" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><span style="font-size: large;"><img border="0" height="221" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEi1dcjhPIKDYpSEyZgm1SMEa7GkDJgKRCiu2ds6h5wHFbWbgTfK51Crl35qcUosu0a6Wrtt9EVEznD8NRT4XIE_Xd3HaM-PKOWZA2M3dCHLe-XVNYMLs3kGmzi_BRuvCUquN8bQ998Jbvc/s320/1007-5841-brain1_cover.jpg" width="320" /></span></a></div>
<span style="font-family: "helvetica";"><span style="font-size: large;">The first lab I joined (molecular neuroscience) we were researching stroke. Most of the
time I was on a microscope looking at proteins in a rat’s brain. The second lab I joined
(cellular neuroscience) we were researching the properties of single neurons. There I
got the opportunity to look at calcium flowing inside a single neuron in an awake
tadpole. Now (the cognitive neuroscience lab) I spend my time either in an MRI room (a
huge magnet that allows us to image brain activity) or in a computer room handling the
MRI data.
</span></span><br />
<span style="font-family: "helvetica";"><span style="font-size: large;"><br /></span></span>
<span style="font-family: "helvetica";"><span style="font-size: large;">My week days are like: wake up at 6-630, yogurt and coffee, in the lab at 8; crunch
some numbers on MATLAB, crunch some more numbers on python, run down for a
scan at 130, lunch at 330; back in the lab at 430, try to do some more number
crunching, feel exhausted and burnt out, go home; 630 feel bored, download a new
dataset, play around with python, exhausted again 930. Sleep, wake, repeat with a
smile on your face because yours is the best job ever.
</span></span><br />
<span style="font-family: "helvetica";"><span style="font-size: large;"><br /></span></span>
<span style="font-family: "helvetica";"><span style="font-size: large;">Its a lot of hard work. The hardest part of it all is not being able to work for longer hours
because I still haven’t figured out that perfect memory jazz.
</span></span></div>
</div>
<span style="font-size: large;"><img alt="page1image25320" height="0.591797" src="file:///page1image25320" width="141.292997" /></span>
</div>
</div>
Anonymoushttp://www.blogger.com/profile/13266163205297209914noreply@blogger.com0tag:blogger.com,1999:blog-7012561022269203683.post-5132692105885385312016-06-30T16:11:00.002-07:002016-06-30T16:15:23.429-07:00GLM, MVPA, RSA and Null Results<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal;">
<span style="-webkit-font-kerning: none;"><b><span style="font-size: large;"><br /></span></b></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal;">
<span style="font-kerning: none;"><b><span style="font-size: large;">GLM, MVPA, RSA and Null Results</span></b></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal;">
<span style="font-kerning: none;"><span style="font-size: large;">In general linear models (GLMs) one tests the hypothesis, individually for every voxel, that that voxel’s activity is driven by some experimental variable. GLMs allow us to test the hypotheses that each voxel’s time course is that expected by a voxel that responds selectively to each experimental condition. Because this involves testing a null hypotheses for each voxel individually, such a massive univariate analysis is only able to capture changes in the signal-to-noise ratio of individual voxels. </span></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal; min-height: 13px;">
<span style="font-size: large;"><span style="font-kerning: none;"></span><br /></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal;">
<span style="font-kerning: none;"><span style="font-size: large;">Alternatively multivariate approaches, such as mulitvoxel pattern analysis (MVPA), allow us to (i) detect meaningful activity that is distributed across many voxels and (ii) does not rely on a priori specifying a model of the expected voxel time course for each experimental condition. Instead of asking is this voxel (and that voxel, and that other voxel over there, and it’s voxel-neighbour) more active for one condition over the other, MVPA asks: is each condition associated to a particular pattern of activation across (say) 1000 voxels? In other words, we take each voxel to be a feature (dimension, variable) in a high dimensional space and try to model each experimental condition as a linear combination of these features: one kind of activity is observed during one condition, while another kind of activity is present in another condition.</span></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal; min-height: 13px;">
<span style="font-size: large;"><span style="font-kerning: none;"></span><br /></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal;">
<span style="font-kerning: none;"><span style="font-size: large;">Representational similarity analysis (RSA) is also a multivariate approach, but goes further than MVPA. MVPA asks the yes/no question of whether the activity in a group of voxels for one experimental condition is different enough from the other condition (different enough that we could decode condition given activity). RSA goes further in quantifying <i>how </i>different the activity patterns are. </span></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal; min-height: 13px;">
<span style="font-size: large;"><span style="font-kerning: none;"></span><br /></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal;">
<span style="font-kerning: none;"><span style="font-size: large;">In order to better understand RSA, I ran an RSA analysis on a freely available fMRI dataset <a href="https://openfmri.org/dataset/ds000157/"><span style="-webkit-font-kerning: none;">https://openfmri.org/dataset/ds000157/</span></a>. In this study participants were scanned while seeing pictures of either food or tools. The experiment, a block design, consisted of showing each participant 24s of pictures of food, followed by 24s of pictures of tools, followed by 24s of food, followed by 24 of tools…. and so on 8 times for each condition.</span></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal; min-height: 13px;">
<span style="font-size: large;"><span style="font-kerning: none;"></span><br /></span></div>
<br />
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal;">
<span style="-webkit-font-kerning: none;"><span style="font-size: large;">The first thing I did was to attempt to replicate the original analysis and run a GLM to ask: which voxels are more active for food compared to tools?</span></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal;">
<span style="-webkit-font-kerning: none;"><span style="font-size: large;"><br /></span></span></div>
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEiHcocR5ziq_AKPGwET2ijjShW6paQY4nun3v6yu6RMjwK5Pp5sL6eMQiMCfjOYjC9213kuwBTXmHf7Umtw_i19tTcFcbK-avf8mHdwR2dA7fmDfo_-n_KnBb6rOOUk14WlAgFXnSPqY_0/s1600/Screen+Shot+2016-06-30+at+19.08.34.png" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><span style="font-size: large;"><img border="0" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEiHcocR5ziq_AKPGwET2ijjShW6paQY4nun3v6yu6RMjwK5Pp5sL6eMQiMCfjOYjC9213kuwBTXmHf7Umtw_i19tTcFcbK-avf8mHdwR2dA7fmDfo_-n_KnBb6rOOUk14WlAgFXnSPqY_0/s1600/Screen+Shot+2016-06-30+at+19.08.34.png" /></span></a></div>
<div class="separator" style="clear: both; text-align: center;">
<span style="font-size: large;"><br /></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal;">
<span style="-webkit-font-kerning: none;"><span style="font-size: large;">On the figure above, to the left are the results of the original experiment reported in <a href="http://www.ncbi.nlm.nih.gov/pubmed/23578759"><span style="-webkit-font-kerning: none;">http://www.ncbi.nlm.nih.gov/pubmed/23578759</span></a>. To the right, are the results of my GLM (second level FWE p<.05). Kinda different… Although we both found activity in the visual cortex and insula, I found two distinct insula clusters (k=5 each). </span></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal; min-height: 13px;">
<span style="font-size: large;"><span style="font-kerning: none;"></span><br /></span></div>
<div class="separator" style="clear: both; text-align: center;">
</div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal;">
<span style="font-kerning: none;"><span style="font-size: large;">So in an attempt to validate my own results, to try to contribute to science and to kill some time, I ran an RSA. First, I ran another GLM without spatial smoothing and where each block was modelled separately. Then, I used the SPM toolbox MarsBar to extract the voxels that were active according to my GLM, resulting in 5 ROI masks: L and R visual (k>100 each), left superior and left inferior insula (k=5 each) and another cluster (not shown in the figure above) somewhere between the parahippocampal region and the amygdala (k=17). Then, for each participant, for each ROI, I constructed a similarity matrix. Each entry in these matrices is the correlation coefficient for the betas estimated for every pair of condition. Then, for each ROI, I took the average of that ROI’s similarity matrix across participants, giving me the following:</span></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal;">
<span style="-webkit-font-kerning: none;"><span style="font-size: large;"><br /></span></span></div>
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEj_uTjuPaNC8w_pstWJiXbjjt7PHHqU0snD1LHGjoNs-sbTq3C6YzJPCkIx_3uRM7WeLcQor-9PoSERmSfnXAslBes2JwMkyagKpcXQPK0pkaG7wNk74gePYi-hqeBKx0yPduetYP-KeRI/s1600/Screen+Shot+2016-06-30+at+19.08.44.png" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><span style="font-size: large;"><img border="0" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEj_uTjuPaNC8w_pstWJiXbjjt7PHHqU0snD1LHGjoNs-sbTq3C6YzJPCkIx_3uRM7WeLcQor-9PoSERmSfnXAslBes2JwMkyagKpcXQPK0pkaG7wNk74gePYi-hqeBKx0yPduetYP-KeRI/s1600/Screen+Shot+2016-06-30+at+19.08.44.png" /></span></a></div>
<div class="separator" style="clear: both; text-align: center;">
<span style="font-size: large;"><br /></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal;">
<span style="-webkit-font-kerning: none;"><span style="font-size: large;">Pretty no? I think so. But what does it tell us? Unfortunately I’m not quite sure. The insula and parahippocampal/amygdala matrices suggest these regions weren’t really showing similar activity across blocks except, possibly, for neighbouring blocks. Interestingly however, the left superior insula seems to have more structure than the inferior insula, and that structure seems to be closely related to that of the hippocampal/amygdala matrix (although I didn’t quantify this similarity - but with RSA I could). The visual cortex matrices seem to keep a consistently high similarity across every condition, but notice the dark upper left square suggesting the first few conditions were especially similar. </span></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal; min-height: 13px;">
<span style="font-size: large;"><span style="font-kerning: none;"></span><br /></span></div>
<div class="separator" style="clear: both; text-align: center;">
</div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal;">
<a href="https://www.blogger.com/blogger.g?blogID=7012561022269203683" imageanchor="1" style="clear: right; float: right; margin-bottom: 1em; margin-left: 1em;"><img border="0" src="webkit-fake-url://7e0e3e19-4291-4337-82b1-5207e7a86a44/image.tiff" style="cursor: move;" /></a><span style="-webkit-font-kerning: none;"><span style="font-size: large;">In the original experiment (and hence in the data available to me) the authors also asked each subject to rate on a scale of 1-5 whether or not they were hungry. They did this before and after scanning. So what I did next (and this is where the real elegance of RSA shines through) was to ask whether I could find a relationship between the reported change in appetite and the change in brain activity. I reasoned: if their reported appetite changed, their brain state changed. RSA is the perfect tool to test this hypothesis because it allows me to actually quantify by how much the activity changed. None of my correlation coefficients were significant. And (as Forrest would put it) that’s what I have to say about that. </span></span><br />
<span style="-webkit-font-kerning: none;"><span style="font-size: large;"><br /></span></span>
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgdL1HZWLPiKVIcHHw2-uYKZ_78Wox_90vYFrvgX1-BEvstoSq-DbrMpfVE_YZKlVV3fP-rmmVi_ZReOLWFbSr_l4NG-gKkeAEfuuwkiSZLJB-H4CLu2XGM51bfBdj28Peee7b4oCUwPhk/s1600/Screen+Shot+2016-06-30+at+19.08.49.png" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgdL1HZWLPiKVIcHHw2-uYKZ_78Wox_90vYFrvgX1-BEvstoSq-DbrMpfVE_YZKlVV3fP-rmmVi_ZReOLWFbSr_l4NG-gKkeAEfuuwkiSZLJB-H4CLu2XGM51bfBdj28Peee7b4oCUwPhk/s1600/Screen+Shot+2016-06-30+at+19.08.49.png" /></a></div>
<span style="-webkit-font-kerning: none;"><span style="font-size: large;"><br /></span></span></div>
<div class="separator" style="clear: both; text-align: center;">
<span style="-webkit-font-kerning: none; margin-left: 1em; margin-right: 1em;"><span style="font-size: large;"></span></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal;">
<span style="font-kerning: none;"><span style="font-size: large;">Looking back, it would have been better if I had ran my RSA GLMs not on the entire brain but only on the ROI voxels. Because it would involve testing less hypotheses (smaller number of voxels) the correction rate would be smaller giving me more power. If I were to get really fancy I could also have used what’s called a searchlight: instead of asking whether the similarity structure in each ROI corresponds to the change in appetite, a searchlight involves asking: where in the brain is a similarity structure observed that matches the change in appetite? This is where RSA becomes extremely powerful, in giving us the ability to map out the representational space by relating similarity in brain activity to similarity in behavioural measures. </span></span></div>
<div class="separator" style="clear: both; text-align: center;">
<span style="-webkit-font-kerning: none; margin-left: 1em; margin-right: 1em;"><span style="font-size: large;">
</span></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal; min-height: 13px;">
<span style="font-size: large;"><span style="font-kerning: none;"></span><br /></span></div>
<div>
</div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal;">
<span style="-webkit-font-kerning: none;"><span style="font-size: large;"><br /></span></span></div>
Anonymoushttp://www.blogger.com/profile/13266163205297209914noreply@blogger.com0tag:blogger.com,1999:blog-7012561022269203683.post-17353669241992425582016-05-28T11:09:00.000-07:002016-05-28T11:09:45.542-07:00An empathy party <div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Arial; line-height: normal;">
<span style="-webkit-text-stroke-color: rgb(64, 64, 64); -webkit-text-stroke-width: initial; color: #404040;"><span style="font-size: large;">In Xanadu did Kubla Khan </span></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Arial; line-height: normal; min-height: 15px;">
<span style="font-size: large;"><span style="-webkit-text-stroke-color: rgb(64, 64, 64); -webkit-text-stroke-width: initial; color: #404040;">A stately pleasure-dome decree: </span><span style="font-kerning: none;"></span></span></div>
<div style="-webkit-text-stroke-color: rgb(64, 64, 64); -webkit-text-stroke-width: initial; color: #404040; font-family: Arial; line-height: normal;">
<span style="font-kerning: none;"><span style="font-size: large;">Where Alph, the sacred river, ran </span></span></div>
<div style="-webkit-text-stroke-color: rgb(64, 64, 64); -webkit-text-stroke-width: initial; color: #404040; font-family: Arial; line-height: normal;">
<span style="font-kerning: none;"><span style="font-size: large;">Through caverns measureless to man </span></span></div>
<div style="-webkit-text-stroke-color: rgb(64, 64, 64); -webkit-text-stroke-width: initial; color: #404040; font-family: Arial; line-height: normal;">
<span style="font-kerning: none;"><span style="font-size: large;"> Down to a sunless sea. </span></span></div>
<div style="-webkit-text-stroke-color: rgb(64, 64, 64); -webkit-text-stroke-width: initial; color: #404040; font-family: Arial; line-height: normal;">
<span style="font-kerning: none;"><span style="font-size: large;"> (S.T. Coleridge)</span></span></div>
<div style="-webkit-text-stroke-color: rgb(64, 64, 64); -webkit-text-stroke-width: initial; color: #404040; font-family: Arial; line-height: normal;">
<span style="font-kerning: none;"><span style="font-size: large;"><br /></span></span></div>
<div style="-webkit-text-stroke-color: rgb(64, 64, 64); -webkit-text-stroke-width: initial; color: #404040; font-family: Arial; line-height: normal;">
<span style="font-kerning: none;"><b style="-webkit-text-stroke-color: rgb(0, 0, 0); color: black;"><span style="font-size: large;">An empathy party </span></b></span></div>
<div style="-webkit-text-stroke-color: rgb(64, 64, 64); -webkit-text-stroke-width: initial; color: #404040; font-family: Arial; line-height: normal; min-height: 15px;">
<span style="font-size: large;"><span style="font-kerning: none;"></span><br /></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Arial; line-height: normal;">
<span style="font-kerning: none;"><span style="font-size: large;">The following reflection occurred yesterfridaynight. The setting: a party; the concoction: gin, then wine, then gin and a rich social setting. <a href="http://brainonbrains.blogspot.ca/2016/04/my-model-of-you-part-i-here-is.html?view=flipcard" target="_blank">As per usual</a> under such circumstances, I was forced into thinking about theory of mind and epistemology. How much / what can people know about each other’s states of mind?</span></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Arial; line-height: normal; min-height: 15px;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEik543TA3YiQuJl1Swxm690erVDZSOO-pSXALF0CtHg7ZfpaqEV1czjNMuh3MziMozbVUm5iFCRlQ_Vpi5vxSC-oQGUB9ofV_D3el4aUa5ZsXGOh9BAlNQ8BtvD0HH2e7BF2Wqc-2b_8Dc/s1600/empathy.jpg" imageanchor="1" style="clear: right; float: right; margin-bottom: 1em; margin-left: 1em;"><span style="font-size: large;"></span></a><span style="font-size: large;"><span style="font-kerning: none;"></span><br /></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Arial; line-height: normal;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEik543TA3YiQuJl1Swxm690erVDZSOO-pSXALF0CtHg7ZfpaqEV1czjNMuh3MziMozbVUm5iFCRlQ_Vpi5vxSC-oQGUB9ofV_D3el4aUa5ZsXGOh9BAlNQ8BtvD0HH2e7BF2Wqc-2b_8Dc/s1600/empathy.jpg" imageanchor="1" style="clear: right; float: right; margin-bottom: 1em; margin-left: 1em;"><img border="0" height="244" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEik543TA3YiQuJl1Swxm690erVDZSOO-pSXALF0CtHg7ZfpaqEV1czjNMuh3MziMozbVUm5iFCRlQ_Vpi5vxSC-oQGUB9ofV_D3el4aUa5ZsXGOh9BAlNQ8BtvD0HH2e7BF2Wqc-2b_8Dc/s320/empathy.jpg" width="320" /></a><span style="font-kerning: none;"><span style="font-size: large;">To clarify, the question does not concern mind reading and coupled dynamics, where the interest lies in behaviour prediction. I think <i>empathy</i> is the phenomenon I am trying to pin down here. Broadly, empathy as I understand it is an emotional state whose evolution (i.e. how it changes over time) depends on the emotional states of others. </span></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Arial; line-height: normal;">
<span style="font-size: large;"><br /></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Arial; line-height: normal;">
<span style="font-size: large;"><span style="font-kerning: none;"></span></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Arial; line-height: normal;">
<span style="-webkit-text-stroke-width: initial;"><span style="font-size: large;">Given my mildly sleep deprived state, and the lasting sooth of endorphins from that afternoons run, my mind was fairly slow. I was more sniffing and feeling out my environment then the usual navigating through it. That is to say, the empathy meter was outpouring like the rivers of Kubla Khan. </span></span></div>
<br />
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Arial; line-height: normal; min-height: 15px;">
<span style="font-size: large;"><span style="font-kerning: none;"></span><br /></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Arial; line-height: normal;">
<span style="font-kerning: none;"><span style="font-size: large;">Among the modern theories that regard empathy, is the <a href="http://www.sciencedirect.com/science/article/pii/S1364661398012625" target="_blank">simulation theory</a>. As the name would suggest, this theory explains the brain’s ability to understand other people’s mental state as a process of simulation. Here is the idea: because brains are similar, they can be in similar states; in empathizing, brains mimic the activity that pertains to the emotional state of another. </span></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Arial; line-height: normal; min-height: 15px;">
<span style="font-size: large;"><span style="font-kerning: none;"></span><br /></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Arial; line-height: normal;">
<span style="font-kerning: none;"><span style="font-size: large;">Within the framework of this theory, lets consider the variables involved in the process of simulation. What are the ideal conditions for simulating? I can think of two right now. (i) The simulator and the simulated should be close enough in state space; and (ii) The simulator should have enough degrees of freedom to bridge the distance in state space.</span></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Arial; line-height: normal; min-height: 15px;">
<span style="font-size: large;"><span style="font-kerning: none;"></span><br /></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Arial; line-height: normal;">
<span style="font-kerning: none;"><span style="font-size: large;">For example, (i) implies that people from different cultures will have a harder time empathizing with each other to the degree that the experiences of emotion in the cultures are different from each other. It also agrees with those situations whereby you and another person just ‘click’. </span></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Arial; line-height: normal; min-height: 15px;">
<span style="font-size: large;"><span style="font-kerning: none;"></span><br /></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Arial; line-height: normal;">
<span style="font-kerning: none;"><span style="font-size: large;">Brains are made of neurons, neurons communicate with one another at certain frequencies (i.e. at a given number of times per second). One of the ways in which we (the cognitive neuroscience community) quantify brain states is by measuring these frequencies. </span></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Arial; line-height: normal; min-height: 15px;">
<span style="font-size: large;"><span style="font-kerning: none;"></span><br /></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Arial; line-height: normal;">
<span style="font-kerning: none;"><span style="font-size: large;">Back to the experience of ‘clicking’ with another person. I would venture a guess that these are cases where the frequencies in brains are similar enough that they resonate. (It would be cool to figure out the equations that dictate this resonance and that predict the frequency bands where this resonance would occur for any two given individuals). (So just to take us out on an even more distant tangent, state space is not linear because you can be quite different from another person but resonance is nearby).</span></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Arial; line-height: normal; min-height: 15px;">
<span style="font-size: large;"><span style="font-kerning: none;"></span><br /></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Arial; line-height: normal;">
<span style="font-kerning: none;"><span style="font-size: large;">Anyway, I digress. Let’s take up (ii) now: the more degrees of freedom (maliability) the simulator has, the larger the gaps in state space it will be able to bridge. This is just a statement about people who are said to be in a highly empathetic state - my case last night.</span></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Arial; line-height: normal; min-height: 15px;">
<span style="font-size: large;"><span style="font-kerning: none;"></span><br /></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Arial; line-height: normal;">
<span style="font-kerning: none;"><span style="font-size: large;">I was sitting on the table, dangling my legs and sipping some gin - just a merrily being a part of the evolution of the social dynamics. All of a sudden I see a cloud come into the room. wtf? It was your facial expression. I look around, sample the scene, venture a hypothesis, you confirm it. I am no longer just merrily being a part of the evolution of the social dynamics. I forget about it. Not really. I detach, go outside. Realize why I went outside. They leave. We resonate. </span></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Arial; line-height: normal; min-height: 15px;">
<span style="font-size: large;"><span style="font-kerning: none;"></span><br /></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Arial; line-height: normal; min-height: 15px;">
<span style="font-kerning: none;"><span style="font-size: large;"> </span></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Arial; line-height: normal; min-height: 15px;">
<span style="font-size: large;"><span style="font-kerning: none;"></span><br /></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal;">
</div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Arial; line-height: normal; min-height: 15px;">
<br /></div>
<div>
<span style="font-kerning: none;"><br /></span></div>
Anonymoushttp://www.blogger.com/profile/13266163205297209914noreply@blogger.com0tag:blogger.com,1999:blog-7012561022269203683.post-42921696908791920862016-05-22T17:57:00.003-07:002016-05-22T17:59:37.957-07:00<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: 'Times New Roman'; line-height: normal;">
<div style="text-align: center;">
<span style="-webkit-font-kerning: none; font-size: large;"><b>On Embodied Cognition and the Information Processing Metaphor: </b></span><br />
<span style="-webkit-font-kerning: none; font-size: large;"><b>a response to “The empty brain”</b></span></div>
<span style="-webkit-font-kerning: none; font-size: large;"><b><br /></b></span>
<span style="-webkit-font-kerning: none; font-size: large;"><b>The empty brain</b></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: 'Times New Roman'; line-height: normal;">
<span style="-webkit-font-kerning: none; font-size: large;"><br /></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: 'Times New Roman'; line-height: normal;">
<span style="-webkit-font-kerning: none; font-size: large;">Robert Epstein recently wrote an article entitled “<a href="https://aeon.co/essays/your-brain-does-not-process-information-and-it-is-not-a-computer" target="_blank">The empty brain</a>.” His position was made clear form the outset: “Your brain does not process information, retrieve knowledge or store memories. In short: your brain is not a computer.” My position, also nice and clear, will be “Hearts pump, kidneys filter, brains … compute” (a quote I attribute to Daniel Dennett).</span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: 'Times New Roman'; line-height: normal; min-height: 15px;">
<span style="font-size: large;"><span style="font-kerning: none;"></span><br /></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: 'Times New Roman'; line-height: normal;">
<span style="font-kerning: none; font-size: large;">This past semester I took a course on cognitive neuroscience. We read two recently published papers a weeks. Every paper had at least one of the following words: “computation”, “information” or “representation.” No doubt, the information processing (IP) metaphor really is pervasive. </span><br />
<span style="font-size: large;"><span style="font-kerning: none;"><br /></span>
</span><br />
<table align="center" cellpadding="0" cellspacing="0" class="tr-caption-container" style="margin-left: auto; margin-right: auto; text-align: center;"><tbody>
<tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjAXzRNRIOBfpKCmU_J3EfRglwvT6qLHRN7kb1REBH8w10Z0NOYmKadUKOkzsZzeWauCAAfiFv3TUG2hD-beRSvoScDlLprFa_uQcG28FJWlh2Uw5mhyphenhyphenN-808iG6DtKgha1BoDDMFzs0sI/s1600/Screen+Shot+2016-05-22+at+17.57.32.png" imageanchor="1" style="margin-left: auto; margin-right: auto;"><span style="font-size: large;"><img border="0" height="150" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjAXzRNRIOBfpKCmU_J3EfRglwvT6qLHRN7kb1REBH8w10Z0NOYmKadUKOkzsZzeWauCAAfiFv3TUG2hD-beRSvoScDlLprFa_uQcG28FJWlh2Uw5mhyphenhyphenN-808iG6DtKgha1BoDDMFzs0sI/s400/Screen+Shot+2016-05-22+at+17.57.32.png" width="400" /></span></a></td></tr>
<tr><td class="tr-caption" style="text-align: center;"><span style="font-size: large;">Taken from DiCarlo's lab <a href="http://dicarlolab.mit.edu/" target="_blank">webpage</a></span></td></tr>
</tbody></table>
</div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: 'Times New Roman'; line-height: normal;">
<span style="font-kerning: none; font-size: large;">After asserting that “the idea that humans must be information processors just because <i>computers</i> are information processors is just plain silly”, Robert rhetorically asks “If the IP metaphor is so silly, why is it so sticky?” My answer, in short, is: because its a good one. </span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: 'Times New Roman'; line-height: normal; min-height: 15px;">
<span style="font-size: large;"><span style="font-kerning: none;"></span><br /></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: 'Times New Roman'; line-height: normal;">
<span style="font-kerning: none; font-size: large;">It seems to me that Robert implicitly assumes that the consequence of the embodied cognition movement is the falsification of the IP metaphor. Below I try to show that the embodied cognition movement and the IP metaphor are not mutually exclusive. </span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: 'Times New Roman'; line-height: normal; min-height: 15px;">
<span style="font-size: large;"><span style="font-kerning: none;"></span><br /></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: 'Times New Roman'; line-height: normal;">
<span style="font-kerning: none; font-size: large;"><b>Embodied Cognition</b></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: 'Times New Roman'; line-height: normal; min-height: 15px;">
<span style="font-size: large;"><span style="font-kerning: none;"></span><br /></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: 'Times New Roman'; line-height: normal;">
<span style="font-kerning: none; font-size: large;">Nervous systems are wonderful things. Each one of us has a network of billions of neural cells with trillions of connections. Day in and day out, for every day of our lives, these very same cells shuffle ions back and forth across their membranes. The result of this ion shuffling? Movement. If there is one thing neurons were ‘made for’, it is navigating that machine we call “body” through space. (“What!?” says the philosopher who uses her brain for thinking about the nature of time). From this realization grew out the movement we now call embodied cognition. </span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: 'Times New Roman'; line-height: normal; min-height: 15px;">
<span style="font-size: large;"><span style="font-kerning: none;"></span><br /></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: 'Times New Roman'; line-height: normal;">
<span style="font-kerning: none; font-size: large;">As Robert puts it, embodied cognition is the thesis that intelligent behaviour is best described as “a direct interaction between organisms and their world.” It is misleading to distinguish between the world as being ‘out there’, and the mind as being ‘in here’, composed of cognitive subsystems, each specialized for a particular task and operating independently of every other. </span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: 'Times New Roman'; line-height: normal; min-height: 15px;">
<span style="font-size: large;"><span style="font-kerning: none;"></span><br /></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: 'Times New Roman'; line-height: normal;">
<span style="font-kerning: none; font-size: large;">To borrow from <a href="https://evanthompson.me/" target="_blank">Evan Thompson</a>, a key figure in the embodied cognition movement, take the example two people dancing together. The situation could be described as two independent individuals acting and reacting to each other. Along these lines, you could give an account of how the visual system of one individual perceives the motions of the other, analyzes the patterns and produces coherent motor outputs of its own. </span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: 'Times New Roman'; line-height: normal; min-height: 15px;">
<span style="font-size: large;"><span style="font-kerning: none;"></span><br /></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: 'Times New Roman'; line-height: normal;">
<span style="font-kerning: none; font-size: large;">Alternatively, you could ‘zoom out’ and describe the two dancers as a single system (a coupled oscillator). Why would you do that? It turns out that this is a better description of how the dance plays out. Social systems (such as the dyad formed by two dancing partners) exhibit <a href="http://onlinelibrary.wiley.com/doi/10.1111/j.1756-8765.2012.01211.x/full" target="_blank">synergistic dynamics</a>. That is, such systems evolve over time according to emergent rules not accounted for in the individual components that constitute it. Simply put, the whole becomes greater than the sum of its parts.</span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: 'Times New Roman'; line-height: normal;">
<span style="font-kerning: none; font-size: large;"><br /></span></div>
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEg3JAOAEMimho9ePBj1M2EMYKT_SxZoKbEqAFQG6EkdSqxmt40pDypgURSOkIE43woK2dJPvTL5tOgtyIIgZgn_rOYwrxPVHt-3sCX5KgufP_sACh1uxX70Ax7bjQyjjYxH9cXZYeDWzmA/s1600/8389005_a-visual-glossary-for-wings-of-wax--dyad_tda25271c.jpg" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><span style="font-size: large;"><img border="0" height="179" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEg3JAOAEMimho9ePBj1M2EMYKT_SxZoKbEqAFQG6EkdSqxmt40pDypgURSOkIE43woK2dJPvTL5tOgtyIIgZgn_rOYwrxPVHt-3sCX5KgufP_sACh1uxX70Ax7bjQyjjYxH9cXZYeDWzmA/s320/8389005_a-visual-glossary-for-wings-of-wax--dyad_tda25271c.jpg" width="320" /></span></a></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: 'Times New Roman'; line-height: normal;">
<span style="font-kerning: none; font-size: large;"><br /></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: 'Times New Roman'; line-height: normal; min-height: 15px;">
<span style="font-size: large;"><span style="font-kerning: none;"></span><br /></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: 'Times New Roman'; line-height: normal;">
<span style="font-kerning: none; font-size: large;">In this description, cognition is not an output of a system called ‘the brain’ that is independent of the environment. Rather, cognition is a property of a system that encompasses both the brain and the environment in which the brain is situated. In the case of two dancers for example, it is not only appropriate to talk about the ‘cognition of the dyad’ above and beyond the ‘cognition of the individuals’, but doing in doing so you can arrive at a better account of how the dance plays out. Does this mean that we should give up on the IP metaphor? I don’t think so.</span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: 'Times New Roman'; line-height: normal; min-height: 15px;">
<span style="font-size: large;"><span style="font-kerning: none;"></span><br /></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: 'Times New Roman'; line-height: normal;">
<span style="font-kerning: none; font-size: large;"><b>Computational Neuroscience </b></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: 'Times New Roman'; line-height: normal; min-height: 15px;">
<span style="font-size: large;"><span style="font-kerning: none;"></span><br /></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: 'Times New Roman'; line-height: normal;">
<span style="font-kerning: none; font-size: large;">As humans we are endowed with a very restrictive cognitive makeup. Our cognitive endowment is far from being at par with the curiosity exemplified by the philosopher of time. (“What <i>is</i> time!” she insists). However, it gives me chills to think of how much we have accomplished given these epistemic limitations. </span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: 'Times New Roman'; line-height: normal; min-height: 15px;">
<span style="font-kerning: none;"><span class="Apple-tab-span" style="font-size: large; white-space: pre;"> </span></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: 'Times New Roman'; line-height: normal;">
<span style="font-kerning: none; font-size: large;">A great deal of our intellectual success hinges on our ability to describe the natural world using formal systems (i.e. mathematics is cool and you should take more math classes). For example, by conceptualizing of neurons as <a href="http://neuronaldynamics.epfl.ch/online/Ch2.S2.html" target="_blank">RC circuits</a> we can use the equations from electrodynamics to describe their firing properties. Doing so advances our understanding of what neurons are and how they do what they do. This process is called scientific model building. </span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: 'Times New Roman'; line-height: normal; min-height: 15px;">
<span style="font-size: large;"><span style="font-kerning: none;"></span><br /></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: 'Times New Roman'; line-height: normal;">
<span style="font-kerning: none; font-size: large;">In model building we take a problem involving a complex system (e.g. how the current of individual ion channels interact to give rise to the firing of an action potential), and simplify it into something that can be manageably described by a small set of equations (e.g. a circuit composed of variable resistors and a capacitor). </span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: 'Times New Roman'; line-height: normal; min-height: 15px;">
<span style="font-size: large;"><span style="font-kerning: none;"></span><br /></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: 'Times New Roman'; line-height: normal;">
<span style="font-kerning: none; font-size: large;">Having a philosophy degree I can’t help but point out that Robert actually contradicts himself. In saying that we do not need the IP metaphor because “all that is required for us to function in the world is for the brain to change in an <i>orderly</i> way”, Robert is verbatim giving us the <a href="https://en.wikipedia.org/wiki/Information_theory#Entropy_of_an_information_source" target="_blank">mathematical definition of information</a>. Information, mathematically described, is a reduction in uncertainty, or entropy. And… what is the opposite of orderly? … Maybe… Chaotically? Maybe… entropically? </span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: 'Times New Roman'; line-height: normal; min-height: 15px;">
<span style="font-size: large;"><span style="font-kerning: none;"></span><br /></span></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: 'Times New Roman'; line-height: normal;">
<span style="font-kerning: none; font-size: large;"><b>The Brain</b></span></div>
<span style="font-size: large;"><br /></span>
<br />
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: 'Times New Roman'; line-height: normal;">
<span style="font-kerning: none; font-size: large;">In short, thinking about the brain in terms of an information processing device is a simplifying assumption (a model) that allows us to formally describe it and in doing so better understand it. Here is my response to Robert’s central claim: Brains aren’t computers any more than neurons are RC circuits, hearts are pumps or kidneys are filters. But thinking of them as such is an extremely useful way of framing our research program. By thinking of it in these terms we can formally describe what kind of thing the brain is and how it evolves. I really don’t think anyone would want to hit a DELETE key on that project.</span></div>
<div>
<span style="font-kerning: none;"><br /></span></div>
Anonymoushttp://www.blogger.com/profile/13266163205297209914noreply@blogger.com0tag:blogger.com,1999:blog-7012561022269203683.post-47417494053078197262016-05-10T19:10:00.002-07:002016-05-10T19:10:50.594-07:00Neurosexy, neurorich and neurogenius?As a young Brazilian highschooler I used to think of neuroscience as that discipline which only complete geniuses could study. The brain is just way too complex. Moreover, in one way or another, the brain is involved in just about every (!) aspect of our lives. This makes neurofindings an especially sexy victim of distorted media claims.<br />
<br />
<a href="http://blogs.discovermagazine.com/neuroskeptic/" target="_blank">Neuroskeptic</a> made a recent video about this very issue.<br />
<div class="separator" style="clear: both; text-align: center;">
<iframe width="320" height="266" class="YOUTUBE-iframe-video" data-thumbnail-src="https://i.ytimg.com/vi/yR2QLKL4FH8/0.jpg" src="https://www.youtube.com/embed/yR2QLKL4FH8?feature=player_embedded" frameborder="0" allowfullscreen></iframe></div>
<br />
As he quite nicely points out, there is copious amounts of neuroscam self-help content on becoming a smarter, sexier and richer you. Quite adeptly neuroskeptic concludes skeptically, saying that in his professional opinion we are not yet living in the age where neuroscience has got it all figured and has got you covered next time you need a few extra bucks or a new date.<br />
<br />
A healthy dose of sober skepticism regarding just about every claim, even if reported in a peer-review article, ... is healthy. After all, as <a href="https://www.youtube.com/watch?v=0Rnq1NpHdmw" target="_blank">John Oliver's</a> really nicely put earlier this week: science is serious shit but even the claims from its participants should be taken with a serious grain of salt.<br />
<br />
With that said, I wrote this blog post to convey my, slightly more optimistic, take on the matters discussed by Neuroskeptic. After all, I too, call myself a(n aspiring) neuroscientist.<br />
<br />
I am a fan of using reading basic science, deriving the implications from the findings and implementing changes to my life accordingly. One immediate example is my sleeping condition. A few months back, the trending neuronews was sleep hygiene: For a good night sleep avoid late night texting and make your room as dark and as quiet as possible. Long story short, I moved my mattress underneath my bed frame and covered the sides making myself a pretty dark and noiseless sleeping dome. Results? Not sure, but based on single subject experimentation (not blinded to condition) I've been sleeping longer hours. This seems to be an example of findings from neuroscience making my life a little bit better.<br />
<br />
Another example I can think of (and this was one of the very first cognitive psychology papers I ever read) concerns encoding. Being a psychology undergraduate I have in the past taken a ridiculous amount of multiple choice exams. How can I spend less time cramming for exams? During my second year I read this paper on how performance on a memory task improves if participants were allowed to sit quietly compared to having to engage in a distractor task. The take home: After flying through a lot of material, don't immediately stand up and go on with your life. Take five minutes to close your eyes, meditate, and allow your brain to encode the information. No data to report here. But this finding, which I read about 4 years ago, is still with me. This seems to be another example of how I lead my life in accordance with what I believe about the brain.<br />
<br />
The above two examples were cases where I used a specific set of findings to make neuroscience-informed life decisions. However, if you spend years studying a particular subject matter you will also develop a keen intuition for the system you are studying (hence the curse of knowledge). Reflecting along these lines, it is inevitable that I see the world through the lens of a neuroscientist. And here, the research I conducted in <a href="http://brainonbrains.blogspot.ca/2016/03/using-machine-learning-to-learn-how.html?view=flipcard" target="_blank">Kurt Haas' lab on neuronal information processing</a>, is especially influential.<br />
<br />
Now, when hanging out, I can't help but know at least something about neurons and how they process information. As such, when watching the world as it unfolds (or more appropriately, when unfolding with the world), this knowledge will tinge my interpretation of just about everything. That is, my knowledge of neuroscience inevitably permeates just about everything I do: from how I interact, to what I think about when I eat oatmeal and yogurt.<br />
<br />
<br />
In short, although I am still a broke student and single it is my professional opinion that my knowledge of neuroscience does allow me to lead a better life.<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />Anonymoushttp://www.blogger.com/profile/13266163205297209914noreply@blogger.com0tag:blogger.com,1999:blog-7012561022269203683.post-47151774423626881512016-04-14T08:36:00.003-07:002016-04-14T08:36:28.230-07:00<div style="text-align: center;">
<b>My Model of You (Part I)</b></div>
<br />
Here is a confession about a personal hobby. Whenever I find myself at a bad party, I diffusely attend to a large group of people and amuse myself by musing about the neural mechanisms underlying social dynamics. It might even involve experimentation! Especially when I have a snap-back on head and wine at hand, I can’t help but perturb comfortable equilibrium states - the evolution of such systems become especially beautiful to watch. However, this begs the (epistemological) question: how much can Andre, the aspiring cognitive neuroscientist, know about the mental states of others from observing behaviour? <br /> It seems obvious that we are all able (to differing degrees) to read facial expressions and interpret gestures. However, everyone has experienced those moments where you hear words, you understand them, but no mental representation of what they might mean comes to mind. On the other hand there are also people (like Sherlock Holmes) who seem to read their environments like a book. In navigating the social environment the brain has to do some heavy lifting. What differs between these two cases? Is there a way to quantify interpersonal decoding? Do the ‘differing degrees’ I hinted at above arise because some brains are better than others at picking up relevant cues or are they better at extracting meaning from the picked up cues?<br /> We could turn to information theory in search for answers to these questions. Originally formulated by Shannon (1948) to quantify the transmission of information across a noisy channel, information theory studies the informational bearing properties of communication systems. Communication involves the transmission of an information bearing signal from sender to receiver. Within this framework, heavily based on probability theory, information is defined as a reduction in uncertainty. <br /> This is important to emphasize because how informative a signal is depends on the degree of uncertainty prior to receiving the signal (sounds bayesian). For example, knowing the outcome of a dice roll is more informative than knowing the outcome of a coin toss because there are six possible outcomes to and only two possible outcomes to a coin. That is, the dice has more uncertainty. Similarly, if someone has a recurring face twitch, their facial gestures will bear little information about what they are currently thinking. On the other hand, someone else who maintains a straight face except when they say ‘peanut butter’, will certainly have something to say about their past experiences with peanut butter. In short, information and degree of randomness go hand in hand.<br /> There is one important dissimilarity between the forms of communication Shannon was studying and interpersonal communication. In conventional communication systems the rules for how to decode a signal are fixed and previously agreed upon. In interpersonal communication however, decoding rules are tacit and fleeting. No one has ever told me what they really mean by the word love. The meaning of signals (such as words or gestures) depend upon a the context in which the interaction occurs. If I tell you I love you after you buy me some ice cream, the word would have a whole other meaning than if I stared right into your eyes and said it with a straight face. Context matters, but why? <br /> The context of a conversation gives rise to the semantic space shared communicating individuals. The process of communication involves probing this space for mutual understanding (Stolk et al., 2016). I emphasize probing because interpersonal communication involves more than just decoding. It involves decoding the decoding rules themselves. More concretely, in social interactions, the meaning of words and gestures depend on the conversational context between the interacting parties. That is, what was just said influences what is meant by what is next said. <br /> This dynamic adjusting of meaning occurs at multiple scales. How you interpret someone else also depends on your shared history with that person. Again, if my mother tells me she loves me I’ll interpret that in one way. If my partner tells me she loves me, I’ll interpret that in quite a different (more lusty kind of) way. The important message here is that meaning depends on context, and context occurs at multiple scales. This dynamic adjustment of decoding rules is what makes interpersonal communication so stimulating for interacting parties and so fascinating for Andre the aspiring cognitive neuroscientist. <br />
<br />
Stolk et al. (2016) Conceptual Alignment: How Brains Achieve Mutual Understanding. <i>TiCS</i>Anonymoushttp://www.blogger.com/profile/13266163205297209914noreply@blogger.com0tag:blogger.com,1999:blog-7012561022269203683.post-52096900377386075882016-03-02T08:23:00.005-08:002016-03-02T08:24:38.302-08:00Using machine learning to learn how neurons learnNeurons compute. That is trivial though, because computers also compute. What is interesting is that neurons learn to compute what they compute. Can't computers also learn? Indeed, there is a a field called machine learning that involves programming a system (e.g. neural network model) to learn from data. Neuronal learning, however, is unparalleled.<br />
<br />
Consider Mrs P. who, after moving to a new country, one foggy night encounters her first cedar tree. From this day on, Mrs P. will effortlessly identify further cedar trees even though she has only experienced a single example under suboptimal conditions. If only we could ask neurons how they learn...<br />
<br />
Wait a minute... we can. It's called science! (jazz hands)*.<br />
<br />
Currently I am interested in what are the general forms of the computations involved in neuronal learning. In machine learning, for example, we use an algorithm called 'back propagation' to train neural network models. What algorithms do <i>real</i> <i>neurons</i> use?<br />
<br />
I have the good fortune to be part of the Haas lab. This means I get to be, right now Weds 2nd March 6am, sitting in front of a computer watching my simplistic 3 layer feed forward neural net fit two-photon data of dendritic calcium traces from neurons before and after they learn (thanks Kasper Podgorski!). This is a blog post though, so I'll break that down.<br />
<br />
We begin by assuming that neurons adjust their computations by changing their functional (firing, conductance) properties. If there is anything inside the cell that we could measure to give us an indication of how these functional properties are changing it is calcium. The problem is that these adjustments in the functional properties of neurons will involve an interplay between many distal regions of the dendritic tree. To this end (and don't ask me how he did it) Kasper built a microscope that detects calcium currents across the dendritic tree. Kasper then went on to image calcium currents in a single neuron inside the intact tadpole while the tadpole learned to detect a stimulus. I now get to play with his dataset : ) Andre is a happy neuroscientist.<br />
<br />
How can I use this dataset to uncover something about the changes in the computations the neuron performs on its inputs during the learning process? Computer modelling! I used machine learning to learn how neurons learn. I reasoned that if I used a data-driven approach to fit two models, one to the calcium currents in the neuron before it learned and another to the post-learning calcium currents, my model parameters would have some information about the computation implemented by the neuron. To that end, I am now training 200 neural network models (123 input units, 1 fully connected hidden layer, 1 output (somatic calcium) layer, 200 iterations of batch BP = 60hours).<br />
<br />
The result of this process is a 200x15,129 dataset (200 networks, 15129 parameters each). The relevant information however, is not in the actual but rather in the relative values of the parameters. I will therefore construct a correlation matrix to get at how these parameters covary across models. My reasoning is that as the neuron hones in on encoding information about a particular stimulus the computation it implements will be 'neater'. That is, I hypothesize that the post-training neuron will have a smaller set of network models that fit it because it implements a more specific computation compared to before learning where it still didn't quite decide what it would be computing. To test that hypothesis I will compare the correlation matrices using some metric that captures how much correlation structure there is overall (currently thinking of using the determinant or the sum of the absolute values, but open to suggestions).<br />
<br />
If my data supports my hypothesis, I have some speculations regarding how the electrophysiological properties of the cell changes during learning (intrinsic plasticity). But that's for another time because I have to do homework - the undergrad life, lol.<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />Anonymoushttp://www.blogger.com/profile/13266163205297209914noreply@blogger.com0tag:blogger.com,1999:blog-7012561022269203683.post-56516714810308643782016-02-21T15:43:00.003-08:002016-02-24T09:31:41.132-08:00Attending to Attention<div style="line-height: normal; text-align: center;">
<div style="text-align: left;">
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal; text-align: center;">
<br /></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal;">
<span class="Apple-tab-span" style="white-space: pre;"> </span>When trail running, I began attending to the different things brain needs to attend to. Consider terrain disambiguation and gait adjustment. Traditionally, object disambiguation and spatial navigation are thought to be distinct operations involving distinct pathways. Intuitively at least, it seems that attending to environment layout and attending to the representation of my body in space are two different processes. Does this suggest there is more than one system of attention? This is unlike because consider how gait can only be adjusted if the terrain is disambiguated and terrain disambiguation is occurring in the context of gaiting behaviour. Therefore it is more likely that attention seems is ubiquitous process rather than being itself discretely deployed computations. I therefore wonder: how is attention instantiated in the brain?</div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal;">
<br /></div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal;">
<span class="Apple-tab-span" style="white-space: pre;"> </span>Rosenberg et al. (2016) propose a neural correlate for attention based on whole brain functional connectivity. Their model, which they call Sustained Attention Network (SAN), was arrived at using a data driven approach. They first had n=25 individuals perform an attention task under an fMRI. Task performance was indexed using d’. Functional network connectivity was operationalized as the matrix of correlation coefficients of activity between brain regions. They then used robust regression to arrive at a model that predicts task performance from network edges (the correlation matrix entries). To evaluate the resulting regression model, they used one-out cross validation: the above procedure was performed n times using the data of n-1 participants, and the arrived at model was used to predict the task performance of the left out individual. The observed and predicted task performance significantly correlated (<i>r </i>= .86) thereby confirming the success of their model in operationalizing attention. Notably Rosenberg et al. (2016) used the SAN model, arrived at by having a western sample perform a standard lab attention task, to successfully predict ADHD symptom severity in a dataset form Peking University. </div>
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal;">
<br /></div>
<br />
<div style="-webkit-text-stroke-color: rgb(0, 0, 0); -webkit-text-stroke-width: initial; font-family: Helvetica; line-height: normal;">
<span class="Apple-tab-span" style="white-space: pre;"> </span>I began with the concern that attention seems to be at once both ubiquitously involved across different cognitive computations and is therefore unlikely to be itself a discrete cognitive process; and the observation that there seems to be a distinct process we call “attention” that is involved in all of the instances in which we use the verb “attend”. Three points are of notable interest. First, these findings suggests the many facets of the mind we denominate attention might best be explained as a distributed network phenomenon. Second, if that is the case, then is it at odds with our subjective experience of attention being a unified process of the mind? What does this tell us about the relationship between our cognitive make up (i.e. the computations implemented by our brains) and our experiences of these processes? Third, traditionally we distinguish between processes such as working memory, perception, attention, etc. Granted these distinctions are in place to make research possible, but is the mind really a bundle of discretely modular processes? It seems to me that as our statistical models further develop from uni- to multivariate, conceptualizing the mind not as many discrete processes but a network of interacting computations will be a more fruitful methodological approach. </div>
</div>
</div>
Anonymoushttp://www.blogger.com/profile/13266163205297209914noreply@blogger.com0