Professor at Stanford. Psychology/Neuroscience/Data Science/AI. Books include: The New Mind Readers, Handbook of fMRI Data Analysis, Hard to Break, Statistical Thinking, and Better Code, Better Science
https://poldrack.github.io/
Russ Poldrack
Loading...
I assigned random gender/ethnicity labels to scientific abstracts from the literature and then asked Claude to do a thematic analysis. Claude identified a clinical versus computational split for female/male authors and a DEI focus for Black/URM authors. All in completely random data.
Estimating parameters: Closed-form and Bayesian estimation
russpoldrack.substack.com/p/estimating... - the latest in my Better Code, Better Science series
Sasha Gusev
Russ Poldrack
It was fun to participate in this! huggingface.co/papers/2606....
Estimating parameters using optimization russpoldrack.substack.com/p/estimating... - the latest in my Better Code, Better Science series
Simulation-based inference russpoldrack.substack.com/p/simulation... - the latest in my Better Code, Better Science series
This is an essential question right now, and @tanialombrozo.bsky.social et al. do a great job of addressing it!
wow, Fable is basically useless for anything related to neuroscience. e.g., in response to the prompt "What is the neural basis of cognitive control?" it falls back to Opus with the attached message. I appreciate guardrails but this makes the model useless for science.
The Stanford Report did a nice writeup of the work that our center has been doing alongside Steve Goodman's SPORR. I think it's a great example of how people can work in tandem on similar topics within an institution and jointly get credit, rather than competing.
Simulating data russpoldrack.substack.com/p/simulating... The latest from my Better Code, Better Science series
"Nobody needs convincing that a science of the Meaning of Life is the wrong project. I will argue that a science of
phenomenal consciousness is wrong for exactly the same reason: neither is amenable to the third-person perspective science requires." by @profdata.bsky.social arxiv.org/abs/2606.00226
Russ Poldrack
Russ Poldrack
Russ Poldrack
Russ Poldrack
Russ Poldrack
Science is constitutively third-personal: its findings are in principle reproducible by any observer, independent of perspective, and answerable to measurement. This is the source of its power and als...
Nice write-up of Stanford's CORES and SPORR groups. Very lucky to be part of a great community of folks (including @russpoldrack.org, Steve Goodman, & @mcxfrank.bsky.social, all mentioned here) working to improve how we science
news.stanford.edu
CORES and SPORR together extend a long Stanford lineage of reliable science.
Do LLMs *understand* language? Do educational AI agents *understand* the material they teach (or their students)? Claims about what AI systems do or don't understand are pervasive, but assessing them requires an account of MACHINE UNDERSTANDING
