Sean Trott
@seantrott.bsky.social
October 10, 2025 at 12:55 AM
Thank you! I'd also recommend this philosophy-oriented overview of "interventionist" methods for studying neural networks: philpapers.org/rec/MILIMF-2
Raphaël Millière & Cameron Buckner, Interventionist Methods for Interpreting Deep Neural Networks - PhilPapers
Recent breakthroughs in artificial intelligence have primarily resulted from training deep neural networks (DNNs) with vast numbers of adjustable parameters on enormous datasets. Due to their complex ...
philpapers.org
October 10, 2025 at 12:53 AM
Thank you! I'd also recommend this philosophy-oriented overview of "interventionist" methods for studying neural networks: philpapers.org/rec/MILIMF-2
Reposted by Sean Trott
Please read my essay in TIME, which @science.org
did not do carefully before publishing this assertion. time.com/7285045/resi...
did not do carefully before publishing this assertion. time.com/7285045/resi...
Why I’m Resigning from the NSF and Library of Congress
I cannot participate in systems that require dishonesty as the price of belonging.
time.com
May 14, 2025 at 4:23 PM
Please read my essay in TIME, which @science.org
did not do carefully before publishing this assertion. time.com/7285045/resi...
did not do carefully before publishing this assertion. time.com/7285045/resi...
In case you're interested I make a similar argument in this blog post, that construct validity (and external validity) are both key problems facing the scientific study of LLMs: seantrott.substack.com/p/llm-ology-...
LLM-ology: the challenges ahead
Three hard questions for a new paradigm.
seantrott.substack.com
May 4, 2025 at 7:31 PM
In case you're interested I make a similar argument in this blog post, that construct validity (and external validity) are both key problems facing the scientific study of LLMs: seantrott.substack.com/p/llm-ology-...
This is great—there's definitely a measurement problem in much (maybe all) of benchmarking.
May 4, 2025 at 7:30 PM
This is great—there's definitely a measurement problem in much (maybe all) of benchmarking.