29 citations · 44 across the 5 of their papers we have counts for
5 papers
Automated Feature Labeling with Token-Space Gradient Descent
Julian Schulz, Seamus Fallows
We present a novel approach to feature labeling using gradient descent in token-space. While existing methods typically use language models to generate hypotheses about feature mea…
Comparing Optimization Targets for Contrast-Consistent Search
Hugo Fry, Seamus Fallows, Ian Fan +2
We investigate the optimization target of Contrast-Consistent Search (CCS), which aims to recover the internal representations of truth of a large language model. We present a new…
Constraints on cosmologies inside black holes
Seamus Fallows, Simon F. Ross
We study the construction of holographic models with closed FRW cosmologies on the worldvolume of a constant-tension brane inside a Schwarzschild-AdS black hole. In dimensions d>2,…
Islands and mixed states in closed universes
Seamus Fallows, Simon F. Ross
We investigate the appearance of islands when a closed universe with gravity is entangled with a non-gravitating quantum system. We use braneworlds in three-dimensional multibounda…
Making near-extremal wormholes traversable
Seamus Fallows, Simon F. Ross
We construct a traversable wormhole from a charged AdS black hole by adding a coupling between the two boundary theories. We investigate how the effect of this deformation behaves…