Time for some major personal updates. I'm delighted to announce that I graduated! Next up: I'll be joining UCLA as an Assistant Professor in January! Please reach out if you're interested in joining the Coalas lab (
@CoalasLab
) as a PhD student or Postdoc 🧠
I'm very happy that I could give my first talk to the CA Meeting on Psycholinguistics
#CAMP3
community who provided extremely useful feedback and comments. I presented empirical evidence that aligns with the hypothesis that production expectations modulate contrastive inference.
Communicating uncertainty is crucial in news, e.g., is the suspect definitely or only maybe guilty? Determining a reader's guilt perception from news can advance effective communication - but it's hard.
We introduce SuspectGuilt: a richly annotated corpus of crime reports.
Excited for this tomorrow (Thursday)! Tune in to hear about a production-centric approach to explain variation in contrastive inference (and/or to see funny objects).
Also -- play with our data yourself here:
Language&Groups; Track4; ID 2433
#cogsci2020
Now I arrived & I'm very excited to start my PhD in
@Stanford
's linguistics department this autumn. I can't wait to learn, work, struggle and celebrate at this amazing place for the next five years. On top of that I'm incredibly happy that I'll be able to join
@ALPSLabStanford
.
This song came out in my first year of CogSci undergrad and since then I've regularly been coming back to it. What a great way to advocate Cognitive Science to a broader audience. Kudos to
@kristian_tylen
and everyone else involved
@AarhusUni
.
Open a new tab and check out our work on how the Rational Speech-Act framework explains contrastive inference variation (with color adjectives)! Poster is already up () -- video will follow soon. Excited to present on Saturday and chat!
#CUNY2020
Tune in Sat to see
@mhahn29
's talk on joint work with
@rljfutrell
and
@LanguageMIT
explaining structural forgetting effects within a lossy context surprisal framework; and visit
@ElisaKreiss
's Fri poster to see RSA explain contrastive inference variation with color adjectives!
@linguistsherry
I made the same experience. Some MTurk workers really take some time to give valuable feedback in the comments or just share their thoughts and wishes. It always makes me happy when going through my data.
Bayesian regression modeling (for factorial designs): A tutorial
@TimoRoettger
and I built on
@BodoWinter
's tutorials to showcase the look-and-feel of a Bayesian regression analysis, using
@paulbuerkner
's
#brms
package. Curious? -> Explore!
Since the image alt text isn't working, I'm adding it here: The first picture shows me and my advisor Chris in the traditional US-American graduation robes. We're standing outside on Stanford campus and are smiling. The second picture shows my yellow UCLA hoodie and small mascot.
Ever had an idea you were SO excited about but just couldn't get to *work*? Do you often daydream about it wistfully, and wonder what could have been? Consider submitting it to our
@NeurIPSConf
workshop "I Can't Believe It's Not Better!":
This work is so cool! TL;DR: If you're trying to assess image alttext quality for BLV users, you can't rely on a context-free method like clipscore because it ignores the communication context/goal :-)
they release BLV+sighted alttext assessments !
And another example of the great job the
#CUNY2020
organizers have done to keep this conference going even virtually.
I'm also presenting and online now, and excited to talk about contrastive inference and weirdly colored fruits: .
I was honored to be asked to give a talk at my old school on Cognitive Science and in particular on studying Cognitive Science at
@UniOsnabrueck
. Interesting questions, fruitful discussions and hopefully new Cognitive Science enthusiasts.
@hawkrobe
@adelegoldberg1
I couldn't agree more. Very excited for your (and everyone else's) talk -- and congrats on winning (!!!) the language computational modeling prize! 🥳🎉
@AndrewLampinen
That's a cool connection! The insertion of "shoulders" could definitely de-noise the interpretation since one usually doesn't sit on the "shoulders" of a horse. "Back" can apply to both but might be more frequent with people and therefore help as well.
What do people use here as pdf editors/annotators? I find Preview a bit cumbersome in the way highlighting and adding notes works. I then switched to Highlights but there are just too many bugs especially when editing old/scanned PDFs. I mainly need it for "active" paper reading.
Video + transcript is up now. I'm excited to chat with you all and hear your thoughts and questions tomorrow (Saturday) during poster session C -- Zoom link is in the OSF project description.
#CUNY2020
@tallinzen
I’ve been wondering about that, too. We’re talking about an age range in which people have the right to vote, drive cars, even buy weapons at some places. This suggests to me that undergrads should be treated and judged (but also excused and heard) like any other adult.
@skeptikantin
@garicgymro
@adelegoldberg1
I think a separate issue here is that audience members don't know whether the speaker is looking at them and don't backchannel through their facial expressions whether they understood, are listening,... I find that almost more off-putting than not seeing them.
@LizBonawitz
@TomerUllman
I share this intuition.I could also imagine that it originated from sundials, where the measurement is not very accurate.Then you would just differentiate 9, 10 or “in between”. Starting from this “half” might have remained as reference as the point being farthest away from 9&10.
@Dr_Semantic
That sounds fun! One take: Possibly because asking a question not only depends on my desire to ask but also on the situational appropriateness,... which could have passed? Maybe some politeness/permission issue as well? (Similar to “Can you give me” not being about ability.)
Stay strong, German negotiators.
According to
@JosephJEsposito
, "any deal Elsevier does with them becomes the de facto deal for the entire world."
- So let's hope.
@TomerUllman
Very cool. Is it a reoccurring pattern that the second output has these repetitions like “complex, complex” and “huge, huge” in the beginning?
@AliyahReh
@eolasinntinn
As an alternative (especially for collaborative projects) I learned to appreciate the “here” package. It might be worth a note, because it potentially makes files a lot cleaner.
@masterwahnsinn
Have a nice trip! It’s so sad that the weather in California isn’t much different from London right now (and it’s been like this for weeks)...
Shout out to everyone with just a bit of lawn. We can all play our part to preserve the insect population — simple actions like these can already make a difference.
@meanwhileina
Same.. and one variation of it: I will never know how all of the stories will end. How long will humanity survive? What will destroy us? Will we make it to other planets? Will we be able to create true artificial intelligence? ...
I had 2 kids under 5 when I did my PhD. This would have been an amazing help! Available for German researchers; needs to be rolled out across the world.