-
Notifications
You must be signed in to change notification settings - Fork 0
/
index.html
480 lines (359 loc) · 16.6 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
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
<!DOCTYPE html>
<html>
<head>
<title>CogBench</title>
<link rel="icon" type="image/x-icon" href="static/images/favicon.ico" />
<link
href="https://fonts.googleapis.com/css?family=Google+Sans|Noto+Sans|Castoro"
rel="stylesheet"
/>
<link rel="stylesheet" href="static/css/bulma.min.css" />
<link rel="stylesheet" href="static/css/bulma-carousel.min.css" />
<link rel="stylesheet" href="static/css/bulma-slider.min.css" />
<link rel="stylesheet" href="static/css/fontawesome.all.min.css" />
<link
rel="stylesheet"
href="https://cdn.jsdelivr.net/gh/jpswalsh/academicons@1/css/academicons.min.css"
/>
<link rel="stylesheet" href="static/css/index.css" />
<script src="https://ajax.googleapis.com/ajax/libs/jquery/3.5.1/jquery.min.js"></script>
<script src="https://documentcloud.adobe.com/view-sdk/main.js"></script>
<script defer src="static/js/fontawesome.all.min.js"></script>
<script src="static/js/bulma-carousel.min.js"></script>
<script src="static/js/bulma-slider.min.js"></script>
<script src="static/js/index.js"></script>
<meta name="viewport" content="width=device-width, initial-scale=1">
</head>
<body>
<!-- <section class="hero is-warning">
<div class="hero-body">
<p class="title">
Under Construction
</p>
<p class="subtitle">
Internal preview only
</p>
</div>
</section> -->
<section class="hero">
<div class="hero-body">
<div class="container is-max-desktop">
<div class="columns is-centered">
<div class="column has-text-centered">
<h1 class="title is-2 publication-title">
CogBench
</h1>
<h1 class="subtitle is-2 publication-title">
Thinking like a human, thinking like a machine:<br/> What's the difference?
</h1>
<div class="is-size-5 publication-authors">
<div class="is-size-5 publication-authors">
<a href="https://minnesotanlp.github.io/">University of Minnesota</a>,
<a href="">University of Wisconsin</a>,
<a href="">Hamline University</a>
<span class="eql-cntrb"
>
<!--
<small><br /><sup>*</sup>Main Developers <sup>+</sup>Contributors</small
></span > -->
</div>
<div class="column has-text-centered">
<div class="publication-links">
<!-- Github link -->
<span class="link-block">
<a
href="https://docs.google.com/forms/d/e/1FAIpQLSdAyv3ELW09sRa1L_xQrDXS9q1aeWtntdjAFpeH4SXJoe6FVQ/viewform"
target="_blank"
class="external-link button is-normal is-rounded is-dark is-outlined"
>
<span>Collaboration Interest Form</span>
</a>
</span>
<!-- Github link -->
<span class="link-block">
<a
href="https://cogbench-prd.ngrok.dev"
target="_blank"
class="external-link button is-normal is-rounded is-dark is-outlined"
>
<span>Data Collection Tool</span>
</a>
</span>
<!-- Github link -->
<span class="link-block">
<a
href="static/pdfs/instructions.pdf"
target="_blank"
class="external-link button is-normal is-rounded is-dark is-outlined"
>
<span>Instructions for Contributors</span>
</a>
</span>
</div>
</div>
</div>
</div>
</div>
</div>
</section>
<section class="section hero is-small left-aligned is-light" style="text-align: left;">
<div class="container is-max-desktop" style="text-align: left;">
<div class="content" style="text-align: left;">
<h3>CogBench is a project bridging machine learning and human-centered disciplines.
<br/><br/>We aim to broadly study cognition across both language models and humans.</h3>
</div>
</div>
</section>
<!-- Research question start -->
<section class="section hero is-small left-aligned" style="text-align: left;">
<div class="container is-max-desktop" style="text-align: left;">
<div class="content" style="text-align: left;">
<h2 class="title is-3"> Members</h2>
We are an interdisciplinary group studying cognition from various research fields, such as Psychology, English, Educational Psychology, and Computer Science.
<br><br>
<p>
<b>Student Members:</b>
<a href="https://karinjd.github.io/" class="link">Karin de Langis*</a>,
<a >Jong Inn Park</a>,
<a >Bin Hu</a>,
<a >Stuti Shah</a>
</p>
<p> <b>Contributors: </b>
<a >Andreas Schramm</a> (Lingusitics),
<a href="https://cla.umn.edu/about/directory/profile/elfen001" >Andrew Elfenbein</a> (English),
<a href="https://sites.google.com/site/michaelmensink/">Mike Mensink</a> (Psychology)
</p>
<p> <b>Principal Investigator: </b><a href="https://dykang.github.io/" >Dongyeop Kang</a> </p>
</div>
</section>
<!-- Research question end -->
<!-- Takeaway start -->
<span id="takeaway">
<!-- The content in this section will be replaced with the content in takeaway-mobile.html if the browser's width is less than 1024px. -->
<section class="section hero is-small is-light">
<div class="container is-max-desktop">
<div class="content">
<h2 class="title is-3 ">Objective</h2>
<p>The project’s goals include:
<ol>
<li>Identify and formalize cognitive and linguistic distinctions between language models and humans</li>
<li> Test theories/models/frameworks developed in human studies within an artificial intelligence setting
</li>
<li> Build foundation for understanding 'internals' of LLM cognition - more predictive power, interpretability for LLM behavior
</li>
<li>Explore feasibility of LLM assistance in materials development and validation for human studies</li>
</ol>
<br/><br/>
</p>
<!--<p>The state of the art in LLMs has advanced rapidly, demanding new suites of benchmarks to assess emergent capabilities. For example, contemporary work explores benchmarking LLMs' Emotional Intelligence (EQ)
and decision making (CogBench). We want to develop a broad spectrum cognitive benchmark that evaluates LLMs across areas of cognition that are well-studied in humans - e.g. selective attention, working memory,
psycholinguistics. Our aim is to understand the ways in which LLM cognition differs from humans, as well as the ways in which it is similar.
</p>
<p><b> Project Goals: </b></p>
<ol>
<li> Create a detailed cognitive inventory to benchmark LLM capabilities</li>
<li> Utilize domain experts to probe cognition in LLMs: Cognitive Science, Psychology, Linguistics
</li>
<li>Facilitate deeper understanding of LLM cognition compared to human cognition</li>
<li> Build foundation for understanding 'internals' of LLM cognition - more predictive power, interpretability for LLM behavior
</li>
</ol>-->
</div>
</div>
</div>
</div>
</ul>
</div>
</div>
</section>
</span>
<!-- Takeaway end -->
<!-- Research question start -->
<section class="section hero is-small">
<div class="container is-max-desktop">
<div class="content">
<h2 class="title is-3">Join our project!</h2>
<div class="level-set has-text-justified">
We are currently expanding our team to include more researchers in psychology, linguistics, cognitive science, and related fields. If you are interested in collaborating and co-authoring with us, please be in touch.
<p> The best way to contact us is by filling out the <a href="https://docs.google.com/forms/d/e/1FAIpQLSdAyv3ELW09sRa1L_xQrDXS9q1aeWtntdjAFpeH4SXJoe6FVQ/viewform" target="_blank" class="external-link is-normal"><span> <u>collaboration interest form</u></span></a>. </p>
<p> We are looking for collaborators willing to contribute data, provide feedback on our data annotation tool, and/or co-author our paper. </p>
</div>
</div>
</section>
<!-- Research question end -->
<!-- Data type start -->
<section class="section hero is-small is-light">
<div class="container is-max-desktop">
<div class="content">
<h2 class="title is-3 ">Demo Video</h2>
<video poster="" id="tree" autoplay controls muted loop height="100%">
<!-- Your video here -->
<source src="static/videos/cogbench-demo-final.mov"
type="video/mp4">
</video>
<p class="subtitle has-text-centered">
View this video for a demo of the data collection web interface. We will provide the login credentials once you fill out our interest form!
</p>
<p><a href="static/videos/cogbench-demo-final.mov" target="_blank" class="external-link is-normal"><span> <b>Link to Video</b></span></a> </p>
</div>
</div>
</div>
</section>
<!-- Data type end -->
<!-- Research Contributions start
<section class="section hero is-small is-light">
<div class="container is-max-desktop">
<div class="columns is-centered">
<div class="column is-full">
<div class="content">
<h2 class="title is-3">Example Output Data</h2>
<div class="level-set has-text-justified">
<img src="static/images/Prompting.png" style="margin-inline: 5px;">
</div>
</div>
</div>
</section>
Research Contributions end -->
<!-- Image carousel -->
<section class="section hero is-small">
<div class="hero-body">
<div class="container is-max-desktop">
<div class="columns is-centered">
<div class="column is-full">
<div class="content">
<h2 class="title is-3">Example Output Data</h2>
<div id="results-carousel" class="carousel results-carousel">
<div class="item">
<!-- Your image here -->
<img src="static/images/prompt_1.png" alt="MY ALT TEXT"/>
<h2 class="subtitle has-text-centered">
</h2>
</div>
<div class="item">
<!-- Your image here -->
<img src="static/images/prompt_2.png" alt="MY ALT TEXT"/>
<h2 class="subtitle has-text-centered">
</h2>
</div>
<div class="item">
<!-- Your image here -->
<img src="static/images/prompt_3.png" alt="MY ALT TEXT"/>
<h2 class="subtitle has-text-centered">
</h2>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</section>
<!-- End image carousel -->
<!-- Takeaway start -->
<span id="takeaway">
<!-- The content in this section will be replaced with the content in takeaway-mobile.html if the browser's width is less than 1024px. -->
<section class="section hero is-small is-light">
<div class="container is-max-desktop">
<div class="content">
<h2 class="title is-3">How will data be used?</h2>
<p><b>Curated Dataset Benchmark (V0)</b></p>
<ul>
<li>~100-200 input/outputs categorized by cognitive process</li>
<li>LLM outputs interpreted by experts contributors who have advanced education in relevant field (professors, post docs, PhD candidates)</li>
<li>Strong theory, reliable</li>
</ul>
<p><b>Open Dataset Benchmark (V1)</b></p>
<ul>
<li>~500+ input/output</li>
<li>Data is provided by anyone*</li>
<li>Outputs can be useful to researchers in human-focused fields</li>
<li>Provide basis of cognitive LLM evaluation</li>
</ul>
<p><b>LLM Evaluation </b></p>
<ul>
<li>Compare LLM and human performance</li>
<li>Does model type, size affect different aspects of cognition?</li>
<li>Do different architectures have different cognitive capabilities?</li>
<li> Interpretability/explainability analysis</li>
</ul>
</div>
</div>
</div>
</div>
</ul>
</div>
</div>
</section>
</span>
<!-- Takeaway end -->
<!-- Limitations start
<section class="section hero is-small is-light" style="text-align: left;">
<div class="container is-max-desktop">
<div class="content">
<div class=" custom-text-left">
<h2 class="title is-3">Instructions for Contributors</h2>
<embed src="static/pdfs/instructions.pdf" width="700" height="375" type="application/pdf">
<p><a href="static/pdfs/instructions.pdf" target="_blank" class="external-link is-normal"><span> <b >Link to Instructions</b></span></a> </p>
</div>
</div>
</section>
-->
<!-- Limitations end -->
<!-- Limitations start -->
<section class="section hero is-small " style="text-align: left;">
<div class="container is-max-desktop">
<div class="content">
<div class=" custom-text-left">
<h2 class="title is-3">Primary Contacts</h2>
<p> Reach out if you have any questions, comments, or concerns!</p>
<p> Karin de Langis - <a href=mailto:dento019@umn.edu>dento019@umn.edu</a> </p>
<p> Dr. Dongyeop Kang - <a href=mailto:dongyeop@umn.edu>dongyeop@umn.edu</a></p>
</div>
</div>
</section>
<!-- Limitations end -->
<footer class="footer">
<div class="container">
<div class="columns is-centered">
<div class="column is-8">
<div class="content">
<p>
Copyright @2024 <a href="https://minnesotanlp.github.io/">Minnesota NLP</a> group</a>
</p>
</div>
</div>
</div>
</div>
</footer>
<script>
function loadHtmlWithScripts(targetElemId, filePath) {
return fetch(filePath)
.then((response) => response.text())
.then((data) => {
const targetElem = document.getElementById(targetElemId);
targetElem.innerHTML = data;
const scripts = targetElem.querySelectorAll("script");
scripts.forEach((script) => {
const newScript = document.createElement("script");
if (script.src) {
newScript.src = script.src;
} else {
newScript.textContent = script.textContent;
}
document.body.appendChild(newScript);
newScript.remove();
});
});
}
function loadTakeaway() {
if (window.innerWidth < 1024) {
loadHtmlWithScripts('takeaway', 'takeaway-mobile.html');
}
}
loadTakeaway();
</script>
<!-- Statcounter tracking code -->
<!-- You can add a tracker to track page visits by creating an account at statcounter.com -->
<!-- End of Statcounter Code -->
</body>
</html>