Ryan M. Nefdt
@ryannefdt.bsky.social
Philosopher of science, author of 'Language, science, and structure (OUP 2023) & 'The Philosophy of Theoretical Linguistics' (CUP 2024). Prof @ University of Cape Town & Bristol. Associate Editor Theoretical Linguistics (De Gruyter): ryannefdt.weebly.com
Yes indeed. If I were inclined to anthropomorphise (and I'm not of course!), I would think of them as machine mnemonics
November 24, 2025 at 8:33 AM
Yes indeed. If I were inclined to anthropomorphise (and I'm not of course!), I would think of them as machine mnemonics
This one was set in Cape Town and surprisingly didn't just film in the picturesque spots but went into the real parts too (close to where I grew up).
November 24, 2025 at 8:03 AM
This one was set in Cape Town and surprisingly didn't just film in the picturesque spots but went into the real parts too (close to where I grew up).
Thanks David! We've got a new version of the paper, where we cut to the chase more quickly :) ling.auf.net/lingbuzz/008...
No Such Thing as a General Learner: Language models and their dual optimization - lingbuzz/008355
What role can the otherwise successful Large Language Models (LLMs) play in the understanding of human cognition, and in particular in terms of informing lan-guage acquisition debates? To contribute t...
ling.auf.net
November 24, 2025 at 7:46 AM
Thanks David! We've got a new version of the paper, where we cut to the chase more quickly :) ling.auf.net/lingbuzz/008...
Congrats old friend! I really enjoyed the book
November 19, 2025 at 7:39 AM
Congrats old friend! I really enjoyed the book
Looks good! Can't wait to read it
October 20, 2025 at 5:43 AM
Looks good! Can't wait to read it
All good my friend! Thank you so much for the post!
July 25, 2025 at 9:29 PM
All good my friend! Thank you so much for the post!
Hey Joe! It's this one that's free access for 2 weeks but happy if people check out the other one too :) cup.org/46WiElu
The Philosophy of Linguistics
Cambridge Core - Philosophy of Science - The Philosophy of Linguistics
cup.org
July 25, 2025 at 9:17 PM
Hey Joe! It's this one that's free access for 2 weeks but happy if people check out the other one too :) cup.org/46WiElu
I see your point. I think some structure of could be captured with textual compression. It might not directly map onto a competence grammar but for predictive purposes (Dennett's initial idea behind real patterns) performance is a guide
May 23, 2025 at 2:56 PM
I see your point. I think some structure of could be captured with textual compression. It might not directly map onto a competence grammar but for predictive purposes (Dennett's initial idea behind real patterns) performance is a guide
This one outlines the original idea: arxiv.org/abs/2405.07987. Really makes me more confident that there are real patterns and ML systems can track them to a certain extent. Of course industry practices and tools also add to the potential converges
The Platonic Representation Hypothesis
We argue that representations in AI models, particularly deep networks, are converging. First, we survey many examples of convergence in the literature: over time and across multiple domains, the ways...
arxiv.org
May 23, 2025 at 1:58 PM
This one outlines the original idea: arxiv.org/abs/2405.07987. Really makes me more confident that there are real patterns and ML systems can track them to a certain extent. Of course industry practices and tools also add to the potential converges