Ethan Mollick

emollick.bsky.social

20 May 2025 20:26

Big: The final version of a randomized, controlled World Bank study finds using a GPT-4 tutor with teacher guidance in a six week afterschool program in Nigeria had "more than twice the effect of some of the most effective interventions in education" ("equating to 1.5 to 2 years" of standard school)

Ethan Mollick

emollick.bsky.social

08 May 2025 05:21

Famously, GPT-4o makes up cites to papers (though error rates appear far lower for cites generated by grounded & Deep Research models). How often does it do that? This large-scale study gives us a clear picture. The AI is also biased towards shorter titles & famous papers. arxiv.org/abs/2504.02767

Sung Kim

sungkim.bsky.social

26 April 2025 00:52

Paper2Code A multi-agent LLM framework that transforms machine learning papers into functional code repositories. Paper: arxiv.org/abs/2504.17192 Repo: github.com/going-doer/P...

Gaël Varoquaux

gaelvaroquaux.bsky.social

24 April 2025 05:59

#ICLR2025 Marine Le Morvan presents "Imputation for prediction: beware of diminishing returns": poster Thu 24th arxiv.org/abs/2407.19804 Concludes 6 years of research on prediction with missing values: Imputation is useful but improvements are expensive, while better learners yield easier gains.

Sung Kim

sungkim.bsky.social

22 April 2025 20:00

Comparing Human versus LLM Judges TREC, which is a community of researchers in information retrieval and natural language processing convened by the NIST, found that an independent human judge correlates better with GPT-4o than a human judge.