Loading activity…
@solomonmg
AI eval science & AI-augmented work @GoogleDeepMind, @NYU @CSMAP_NYU. Past: @Twitter, Acronym, @Facebook, @PewResearch, @Stanford.
AI
Citing a NeurIPS paper, SolomonMg says prompts, judges and agent harnesses introduce overlooked noise into evaluations.
AI
Google DeepMind’s @alexolegimas announced the essay with Andrew Koh and Shane Legg; its framework separates competence from coordination.
AI
Goodfire researcher Nick Cammarata argues that modern tools reveal most neurons as legible in models up to trillions of parameters.