Phillip Wadley
@phillipwadley.bsky.social
Part-time linguist 📖
Part-time GM 🧙♂️
Full-time silly goose 🪿
Part-time GM 🧙♂️
Full-time silly goose 🪿
If you're interested in computational metaphor, you need to check this out! Provides commands for lemmatization, Interpretive diversity, and is great for analysising feature emergence.
Developed by J. Nick Reid and Albert Katz.
Wish I knew about this earlier in my PhD!
rdcu.be/eIHM1
Developed by J. Nick Reid and Albert Katz.
Wish I knew about this earlier in my PhD!
rdcu.be/eIHM1
The RK processor: A program for analysing metaphor and word feature-listing data
rdcu.be
September 29, 2025 at 2:25 PM
If you're interested in computational metaphor, you need to check this out! Provides commands for lemmatization, Interpretive diversity, and is great for analysising feature emergence.
Developed by J. Nick Reid and Albert Katz.
Wish I knew about this earlier in my PhD!
rdcu.be/eIHM1
Developed by J. Nick Reid and Albert Katz.
Wish I knew about this earlier in my PhD!
rdcu.be/eIHM1
Reposted by Phillip Wadley
Can large language models stand in for human participants?
Many social scientists seem to think so, and are already using "silicon samples" in research.
One problem: depending on the analytic decisions made, you can basically get these samples to show any effect you want.
THREAD 🧵
Many social scientists seem to think so, and are already using "silicon samples" in research.
One problem: depending on the analytic decisions made, you can basically get these samples to show any effect you want.
THREAD 🧵
The threat of analytic flexibility in using large language models to simulate human data: A call to attention
Social scientists are now using large language models to create "silicon samples" - synthetic datasets intended to stand in for human respondents, aimed at revolutionising human subjects research. How...
arxiv.org
September 18, 2025 at 7:56 AM
Can large language models stand in for human participants?
Many social scientists seem to think so, and are already using "silicon samples" in research.
One problem: depending on the analytic decisions made, you can basically get these samples to show any effect you want.
THREAD 🧵
Many social scientists seem to think so, and are already using "silicon samples" in research.
One problem: depending on the analytic decisions made, you can basically get these samples to show any effect you want.
THREAD 🧵
Shit is getting so bad now that running stats in R has become a soothing activity for me as of late.
September 20, 2025 at 8:01 PM
Shit is getting so bad now that running stats in R has become a soothing activity for me as of late.
Having way too much fun with this word guessing game based on word2vec.
The nifty thing about this is also you can create your own game and see up to 1000 nearest neighbors—a convenient tool for measuring semantic neighborhood density
semantle.com
The nifty thing about this is also you can create your own game and see up to 1000 nearest neighbors—a convenient tool for measuring semantic neighborhood density
semantle.com
Semantle | Daily Word Guessing Game
Semantle is an engaging word-guessing game that challenges you to find the hidden word through semantic clues. Sharpen your linguistic skills!
semantle.com
June 7, 2025 at 12:03 PM
Having way too much fun with this word guessing game based on word2vec.
The nifty thing about this is also you can create your own game and see up to 1000 nearest neighbors—a convenient tool for measuring semantic neighborhood density
semantle.com
The nifty thing about this is also you can create your own game and see up to 1000 nearest neighbors—a convenient tool for measuring semantic neighborhood density
semantle.com