-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathindex.html
executable file
·258 lines (249 loc) · 10.1 KB
/
index.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
<!doctype html>
<html>
<head>
<meta charset="utf-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0, maximum-scale=1.0, user-scalable=no">
<title>reveal.js</title>
<link rel="stylesheet" href="css/reveal.css">
<link rel="stylesheet" href="css/theme/black.css">
<link rel="stylesheet" href="css/custom.css">
<!-- Theme used for syntax highlighting of code -->
<link rel="stylesheet" href="lib/css/zenburn.css">
<!-- Printing and PDF exports -->
<script>
var link = document.createElement( 'link' );
link.rel = 'stylesheet';
link.type = 'text/css';
link.href = window.location.search.match( /print-pdf/gi ) ? 'css/print/pdf.css' : 'css/print/paper.css';
document.getElementsByTagName( 'head' )[0].appendChild( link );
</script>
</head>
<body>
<div class="reveal">
<div class="slides">
<section data-markdown>
##What artificial agents can teach us about social decisions
John Pearson
Biostatistics and Bioinformatics
[pearsonlab.github.io/artificial-agents-social-decisions](https://pearsonlab.github.io/artificial-agents-social-decisions)
</section>
<section>
<img src="https://web.duke.edu/mind/level2/faculty/pearson/assets/images/lab_logo/plab_logo_light.svg" alt="lab logo", style="border: none; background: none; box-shadow: none">
</section>
<section data-background-image="https://cdn.vox-cdn.com/uploads/chorus_image/image/48621959/GettyImages-161070240.0.0.jpg">
<!-- <div style="background-color: rgba(0, 0, 0, 0.75); padding: 10px" class="fragment fade-left">
<ul>
<li>
Large numbers of variables
</li>
<li>
Complex interactions
</li>
<li>
Complexity is the point!
</li>
</ul>
</div> -->
</section>
<section data-background-image="https://web.duke.edu/mind/level2/faculty/pearson/assets/images/sfn2016/chess.jpg">
<aside class="notes" data-markdown>
### Pros:
- Well-studied, normative solutions
### Cons:
- Highly idealized, limited dynamics
- Biologically aligned?
</aside>
</section>
<section data-background-image="https://web.duke.edu/mind/level2/faculty/pearson/assets/images/sfn2016/penalty-shot.jpg">
<aside class="notes" data-markdown>
- Doesn't matter if it's social
- Requires anticipating another agent
- Repeatable, but lots of variation
- Decisions tightly linked to movement
</aside>
</section>
<section>
<div style="float:left; width:25%">
<p></p>
</div>
<div style="float:left; width:25%">
<img src="https://web.duke.edu/mind/level2/faculty/pearson/assets/images/website/kelsey.jpg" style="height: 300px; width: auto">
<p>Kelsey McDonald</p>
</div>
<div style="float:left; width:25%">
<img src="https://scholars.duke.edu/file/i1293252/image_1293252.jpg" style="height: 300px; width: auto">
<p>Scott Huettel</p>
</div>
<div style="float:left; width:25%">
</div>
</section>
<section>
<h3>Penalty Shot</h3>
<video width="70%" align="center" controls autoplay loop src="https://web.duke.edu/mind/level2/faculty/pearson/assets/videos/penaltyshot/concat_trials.mp4" type="video/mp4">
Your browser does not support the video tag.
</video>
<p class="ref"><a href="https://doi.org/10.1101/385195">McDonald, Broderick, Huettel, Pearson</a></p>
<aside class="notes" data-markdown>
- two monkeys, shooter and goalie (shooter recorded)
- controlled by joysticks
- roles rotated, animals know each other
- repeated sessions
- rapidly learned, rich dynamics
</aside>
</section>
<section>
<div style="width: 50%; float: left; text-align: left">
<video autoplay loop src="https://web.duke.edu/mind/level2/faculty/pearson/assets/videos/penaltyshot/concat_trials.mp4" type="video/mp4" id="trace_video">
Your browser does not support the video tag.
</video>
<script>
var vid = document.getElementById("trace_video");
vid.playbackRate = 1.;
</script>
</div>
<div style="width: 50%; float: right; font-size: .9em" >
<h3>How it works</h3>
<ul style="font-size: 0.85">
<li>Human, computer opponents</li>
<li>Incentive compatible</li>
<li>Coevolving strategies</li>
</ul>
</div>
</section>
<section>
<h3>Highly variable strategies</h3>
<img src="https://web.duke.edu/mind/level2/faculty/pearson/assets/images/penaltyshot/kelsey_pk/trajectories_3.svg" style="border: none; background: none; box-shadow: none; width: 50%; height: auto">
<br>
<img src="https://web.duke.edu/mind/level2/faculty/pearson/assets/images/penaltyshot/kelsey_pk/trajectories_4.svg" style="border: none; background: none; box-shadow: none; width:50%; height: auto">
</section>
<section>
<h3>How do we model this?</h3>
$$p(\text{change}) = \Phi(f(s(t), \omega(t)))$$
$$f(s(t), \omega(t)) \sim \mathrm{GP}(0, k)$$
<br>
<br>
<h3>Informally:</h3>
<ul>
<li>strategy depends on game state $s$, opponent $\omega$</li>
<li>$\mathrm{GP}$ is a distribution over <i>functions</i></li>
<li>Bayesian model: we can quantify uncertainty</li>
<li>Joint model: can do "counterfactual" experiments</li>
</ul>
</section>
<section>
<h3>How likely are you to win?</h3>
<video width="70%" align="center" controls autoplay loop src="https://web.duke.edu/mind/level2/faculty/pearson/assets/videos/penaltyshot/concat_EV.mp4" type="video/mp4">
Your browser does not support the video tag.
</video>
</section>
<section>
<h3>How likely are you to change course?</h3>
<video width="70%" align="center" controls autoplay loop src="https://web.duke.edu/mind/level2/faculty/pearson/assets/videos/penaltyshot/concat_switch.mp4" type="video/mp4">
Your browser does not support the video tag.
</video>
</section>
<section>
<h3>Real and predicted change points</h3>
<div style="float:left; width:50%">
<img src="https://web.duke.edu/mind/level2/faculty/pearson/assets/images/penaltyshot/kelsey_pk/realdata_sub3.svg" style="height: 300px; width: auto">
</div>
<div style="float:right; width:50%">
<img src="https://web.duke.edu/mind/level2/faculty/pearson/assets/images/penaltyshot/kelsey_pk/modelprediction_sub3.svg" style="height: 300px; width: auto">
</div>
</section>
<section>
<h3>How opponent-sensitive are you?</h3>
<video width="70%" align="center" controls autoplay loop src="https://web.duke.edu/mind/level2/faculty/pearson/assets/videos/penaltyshot/concat_oppSens.mp4" type="video/mp4">
Your browser does not support the video tag.
</video>
</section>
<section>
<h3>Does identity matter?</h3>
<ul>
<li>Of course!</li>
<li>But context also matters.</li>
<li>Can we separate identity from context?</li>
<li>Major challenge in social neuro/pscyh</li>
</ul>
</section>
<section>
<h3>Does identity matter?</h3>
<img src="https://web.duke.edu/mind/level2/faculty/pearson/assets/images/penaltyshot/kelsey_pk/Fig3A.svg" style="background: white">
</section>
<section>
<h3>Summary</h3>
<ul>
<li>Flexible models let us study more realistic decisions</li>
<li>Efficient models let us make <i>individualized</i> predictions</li>
<li>Accurate models (artificial agents) let us ask "what if" questions, perform synthetic experiments</li>
</ul>
</section>
<section>
<h2>Sponsors</h2>
<div class="wrapper" style="height: 800px">
<img class="grid-img"
src="https://web.duke.edu/mind/level2/faculty/pearson/assets/images/sfn2016/bd2k_logo.png"
style="grid-column:1/3">
<img class="grid-img"
src="https://web.duke.edu/mind/level2/faculty/pearson/assets/images/sfn2016/US-NIH-NIMH-Logo.svg"
style="grid-column:3/5">
<img class="grid-img"
src="https://web.duke.edu/mind/level2/faculty/pearson/assets/images/misc/Nvidia_image_logo.svg"
style="grid-column:1/3; background: white; padding: 15px">
<img class="grid-img"
src="https://web.duke.edu/mind/level2/faculty/pearson/assets/images/misc/Amazon_logo_plain.svg"
style="grid-column:3/5; background: white; padding: 15px">
</div>
</section>
<section>
<div class="wrapper" style="">
<img
class="grid-item"
src="https://web.duke.edu/mind/level2/faculty/pearson/assets/images/website/kelsey.jpg"
style="grid-column: 1/2">
<img
class="grid-item"
src="https://web.duke.edu/mind/level2/faculty/pearson/assets/images/website/sam.jpg"
style="grid-column: 2/3">
<img
class="grid-item"
src="https://pearsonlab.github.io/images/anne.jpg"
style="grid-column: 3/4">
<img
class="grid-item"
src="https://pearsonlab.github.io/images/jack.jpg"
style="grid-column: 4/5">
<img
class="grid-item"
src="https://pearsonlab.github.io/images/pranjal.jpg"
style="grid-column: 1/2">
<img
class="grid-item"
src="https://web.duke.edu/mind/level2/faculty/pearson/assets/images/lab_logo/plab_logo_light.svg"
style="grid-column: 2/4">
</div>
</section>
</div>
</div>
<script src="lib/js/head.min.js"></script>
<script src="js/reveal.js"></script>
<script>
// More info https://github.com/hakimel/reveal.js#configuration
Reveal.initialize({
history: true,
// math: {
// mathjax: 'https://cdn.mathjax.org/mathjax/latest/MathJax.js',
// config: 'TeX-AMS_HTML-full' // See http://docs.mathjax.org/en/latest/config-files.html
// },
// More info https://github.com/hakimel/reveal.js#dependencies
dependencies: [
{ src: 'plugin/markdown/marked.js' },
{ src: 'plugin/markdown/markdown.js' },
{ src: 'plugin/notes/notes.js', async: true },
{ src: 'plugin/math/math.js', async: true },
{ src: 'plugin/highlight/highlight.js', async: true, callback: function() { hljs.initHighlightingOnLoad(); } }
]
});
</script>
</body>
</html>