51 citations · 53 across the 5 of their papers we have counts for
5 papers
Sparse Autoencoders Find Highly Interpretable Features in Language Models
Hoagy Cunningham, Aidan Ewart, Logan Riggs +2
One of the roadblocks to a better understanding of neural networks' internals is \textit{polysemanticity}, where neurons appear to activate in multiple, semantically distinct conte…
Attention-Only Transformers and Implementing MLPs with Attention Heads
Robert Huben, Valerie Morris
The transformer architecture is widely used in machine learning models and consists of two alternating sublayers: attention heads and MLPs. We prove that an MLP neuron can be imple…
Gauge-invariant uniqueness theorems for -graphs
Robert Huben, S. Kaliszewski, Nadia S. Larsen +1
We prove a version of the result in the title that makes use of maximal coactions in the context of discrete groups. Earlier Gauge-Invariant Uniqueness theorems for -algebras…
Gauge-Invariant Uniqueness and Reductions of Ordered Groups
Robert Huben
A reduction of an ordered group to another ordered group is an order homomorphism which maps each interval bijectively onto . We show that if …
Recursive strategy for decomposing Betti tables of complete intersections
Courtney R. Gibbons, Robert Huben, Branden Stone
We introduce a recursive decomposition algorithm for the Betti diagram of a complete intersection using the diagram of a complete intersection defined by a subset of the original g…