activity
20232025
most citedMathematical Models of Computation in Superposition

1 citations · 1 across the 3 of their papers we have counts for

collaborators
Showing cs.LGShow all

5 papers · 1 filter

cs.LG2025

Interpretability in Parameter Space: Minimizing Mechanistic Description Length with Attribution-based Parameter Decomposition

Dan Braun, Lucius Bushnaq, Stefan Heimersheim +2

Mechanistic interpretability aims to understand the internal mechanisms learned by neural networks. Despite recent progress toward this goal, it remains unclear how best to decompo…

cs.LG20241 cited

Mathematical Models of Computation in Superposition

Kaarel Hänni, Jake Mendel, Dmitry Vaintrob +1

Superposition -- when a neural network represents more ``features'' than it has dimensions -- seems to pose a serious challenge to mechanistically interpreting current AI systems.…

cs.LG2024

The Local Interaction Basis: Identifying Computationally-Relevant and Sparsely Interacting Features in Neural Networks

Lucius Bushnaq, Stefan Heimersheim, Nicholas Goldowsky-Dill +7

Mechanistic interpretability aims to understand the behavior of neural networks by reverse-engineering their internal computations. However, current methods struggle to find clear…

cs.LG2024

Using Degeneracy in the Loss Landscape for Mechanistic Interpretability

Lucius Bushnaq, Jake Mendel, Stefan Heimersheim +5

Mechanistic Interpretability aims to reverse engineer the algorithms implemented by neural networks by studying their weights and activations. An obstacle to reverse engineering ne…

cs.LG2023

Dynamical versus Bayesian Phase Transitions in a Toy Model of Superposition

Zhongtian Chen, Edmund Lau, Jake Mendel +2

We investigate phase transitions in a Toy Model of Superposition (TMS) using Singular Learning Theory (SLT). We derive a closed formula for the theoretical loss and, in the case of…