Willing to bet you could massively outperform this model just by fitting a simple discrete time logistic model for survival. The c-statistic was pitiful. This screams "optimising publications for media engagement" to me which is disappointing to say the least.

Show Post

The adjustments are completely unjustified. Frankly they seem clueless about what is required for developing, evaluating, and reporting a prediction model. The tripod+ai checklist is the giveaway. And a clear sign of peer review failure.

Show Post

@richvn.bsky.social might be of interest to @nature.com News?

Show Post

Just to be very explicit, my claim is that the one simplest and best explanation for this, which resolves any apparent strangeness entirely, is that whoever is “speaking” in LLM output is a fictional character, and the real life thing called an LLM is a machine for writing lil stories about that guy

Show Post

I don’t know of any strong evidence one way or another but there’s a very plausible mechanism by which Pangram could be biased against certain varieties of African English (which is NOT the same as being biased against EAL, which I believe it isn’t)

Show Post