471 citations · 528 across the 4 of their papers we have counts for
4 papers
The Capacity for Moral Self-Correction in Large Language Models
Deep Ganguli, Amanda Askell, Nicholas Schiefer +46
We test the hypothesis that language models trained with reinforcement learning from human feedback (RLHF) have the capability to "morally self-correct" -- to avoid producing harmf…
Scattering States in AdS/CFT
A. Liam Fitzpatrick, Jared Kaplan
We show that suitably regulated multi-trace primary states in large N CFTs behave like `in' and `out' scattering states in the flat-space limit of AdS. Their transition matrix elem…
The Sloan Digital Sky Survey-II Supernova Survey: Technical Summary
Joshua A. Frieman, B. Bassett, A. Becker +98
The Sloan Digital Sky Survey-II (SDSS-II) has embarked on a multi-year project to identify and measure light curves for intermediate-redshift (0.05 < z < 0.35) Type Ia supernovae (…
The Plasma Puddle as a Perturbative Black Hole
Clifford Cheung, Jared Kaplan
We argue that the weak coupling regime of a large N gauge theory in the Higgs phase contains black hole-like objects. These so-called ``plasma puddles'' are meta-stable lumps of ho…