Neuron to Graph: Interpreting Language Model Neurons at Scale

1 code implementation • 31 May 2023 • Alex Foote, Neel Nanda, Esben Kran, Ioannis Konstas, Shay Cohen, Fazl Barez

Conventional methods require examination of examples with strong neuron activation and manual identification of patterns to decipher the concepts a neuron responds to.

Language Modelling

Erasure of Unaligned Attributes from Neural Representations

1 code implementation • 6 Feb 2023 • Shun Shao, Yftah Ziser, Shay Cohen

We present the Assignment-Maximization Spectral Attribute removaL (AMSAL) algorithm, which erases information from neural representations when the information to be erased is implicit rather than directly being aligned to each input example.

Attribute

Tensors over Semirings for Latent-Variable Weighted Logic Programs

no code implementations • WS 2020 • Esma Balkir, Daniel Gildea, Shay Cohen

Semiring parsing is an elegant framework for describing parsers by using semiring weighted logic programs.

Latent-Variable PCFGs: Background and Applications

no code implementations • WS 2017 • Shay Cohen

Machine Translation Question Answering +1

Spectral Learning Algorithms for Natural Language Processing

no code implementations • NAACL 2013 • Shay Cohen, Michael Collins, Dean Foster, Karl Stratos, Lyle Ungar

Shared Logistic Normal Distributions for Soft Parameter Tying in Unsupervised Grammar Induction

no code implementations • 1 Jun 2009 • Shay Cohen, Noah A. Smith

We present a family of priors over probabilistic grammar weights, called the shared logistic normal distribution.

Dependency Grammar Induction Unsupervised Dependency Parsing