With classes starting next week, my 8yo stopped by my office to leave some heat for all the college students. No calculators.
Remote
David Bamman
@dbamman@sigmoid.social
mastodon 4.7.3Associate Professor, School of Information, UC Berkeley. Natural language processing, machine learning, computational social science, digital humanities.
619 Followers
576 Following
12 Posts
Joined November 20, 2022
Replying to
Takeaways: open models are good; popular texts are probably not good barometers of model performance; with the bias toward sci-fi/fantasy, we should be thinking about whose narrative experiences are encoded in these models, and how that influences other behaviors. Work with Kent Chang, Mackenzie Cramer and Sandeep Soni. Details on all books here: https://docs.google.com/spreadsheets/d/1jW7EhsNjIGDMoK2JidyDD7UXH9N0NpEJfWFEj05_LC4 6/6
Cooking to Bach’s Mass in B Minor, which continually blows me away as one of the highlights of human creation, especially the 1967 Klemperer version. https://open.spotify.com/track/0acnTnzvpm6MGFbaZmY8LF?si=Mm2x_v2jQgy5O6AMmMGp1g
I declare ChatGPT to be a performance-enhancing substance for the greatest game: Taboo. All wins with it get an asterisk.
Every time I check out a book from the stacks I realize I have absolutely no idea how this marvel of engineering works. This is, like, thousands of books!
Headed to Abu Dhabi now for EMNLP, the Workshop on NLP + Computational Social Science (https://sites.google.com/site/nlpandcss/home/nlp-css-at-emnlp-2022), and the Workshop on Shared Stores and Lessons Learned (https://ssll-emnlp.github.io/schedule.html). Hope to see some of you there! (And if you are there and want to connect, let me know!)
Replying to
How do we know? Fill-in-the-name: "Wow. I sit down, fish the questions from my backpack, and go through them, inwardly cursing [MASK] for not providing me with a brief biography. I know nothing about this man I’m about to interview. He could be ninety or he could be thirty." People can't guess the name; it's not knowledge of English you need, but the underlying content to know that "Kate" is the only right answer. (GPT-4 gets this one right, from 50 Shades of Grey.)
@LauraNelson@mastodon.social Reticulated pythons are a big yes
@agoldst@mastodon.social That's interesting -- thinking about other lexical terms that do that kind of boundary work, "model", "structural"
@agoldst@mastodon.social "'form' is in the typical close reading a very loosely defined term, including almost anything that interests the close reader" -- "form" is the one word whose multiplicity of vague senses I kept thinking about throughout @lucy@sigmoid.social's work on https://arxiv.org/pdf/2212.09676.pdf
@agoldst@mastodon.social Excellent, thank you! I have the start of a summer reading list