Newest Papers
When Models Manipulate Manifolds: The Geometry of a Counting Task
Unknown Author
Transformer Circuits
No abstract available
No ratings yet
View paper →Tracing Attention Computation Through Feature Interactions
Unknown Author
transformer-circuits.pub
No abstract available
No ratings yet
View paper →On the Biology of a Large Language Model
Unknown Author
Transformer Circuits
No abstract available
No ratings yet
View paper →Circuit Tracing: Revealing Computational Graphs in Language Models
Unknown Author
Transformer Circuits
No abstract available
No ratings yet
View paper →Scaling Monosemanticity: Extracting Interpretable Features from Claude 3 Sonnet
Unknown Author
transformer-circuits.pub
No abstract available
No ratings yet
View paper →Towards Monosemanticity: Decomposing Language Models With Dictionary Learning
Unknown Author
transformer-circuits.pub
No abstract available
No ratings yet
View paper →Superposition, Memorization, and Double Descent
Unknown Author
transformer-circuits.pub
No abstract available
No ratings yet
View paper →Toy Models of Superposition
dimensionality
transformer-circuits.pub
No abstract available
No ratings yet
View paper →In-context Learning and Induction Heads
this we mean that they promote sequence completions like
Ghost Blog
No abstract available
No ratings yet
View paper →Correction: Influence of pseudo-Jahn–Teller activity on the singlet–triplet gap of azaphenalenes - Physical Chemistry Chemical Physics (RSC Publishing) DOI:10.1039/D5CP90132D
Atreyee Majumdar, Komal Jindal, Surajit Das, Raghunathan Ramakrishnan
pubs.rsc.org
No abstract available
No ratings yet
View paper →