Picture for John Lafferty

John Lafferty

Disentangling and Integrating Relational and Sensory Information in Transformer Architectures

Add code
May 26, 2024
Viaarxiv icon

Approximation of relation functions and attention mechanisms

Feb 13, 2024
Viaarxiv icon

Relational Convolutional Networks: A framework for learning representations of hierarchical relations

Add code
Oct 05, 2023
Figure 1 for Relational Convolutional Networks: A framework for learning representations of hierarchical relations
Figure 2 for Relational Convolutional Networks: A framework for learning representations of hierarchical relations
Figure 3 for Relational Convolutional Networks: A framework for learning representations of hierarchical relations
Figure 4 for Relational Convolutional Networks: A framework for learning representations of hierarchical relations
Viaarxiv icon

The Relational Bottleneck as an Inductive Bias for Efficient Abstraction

Sep 12, 2023
Figure 1 for The Relational Bottleneck as an Inductive Bias for Efficient Abstraction
Figure 2 for The Relational Bottleneck as an Inductive Bias for Efficient Abstraction
Viaarxiv icon

Abstractors: Transformer Modules for Symbolic Message Passing and Relational Reasoning

Apr 01, 2023
Figure 1 for Abstractors: Transformer Modules for Symbolic Message Passing and Relational Reasoning
Figure 2 for Abstractors: Transformer Modules for Symbolic Message Passing and Relational Reasoning
Figure 3 for Abstractors: Transformer Modules for Symbolic Message Passing and Relational Reasoning
Figure 4 for Abstractors: Transformer Modules for Symbolic Message Passing and Relational Reasoning
Viaarxiv icon

From seeing to remembering: Images with harder-to-reconstruct representations leave stronger memory traces

Add code
Feb 21, 2023
Figure 1 for From seeing to remembering: Images with harder-to-reconstruct representations leave stronger memory traces
Figure 2 for From seeing to remembering: Images with harder-to-reconstruct representations leave stronger memory traces
Figure 3 for From seeing to remembering: Images with harder-to-reconstruct representations leave stronger memory traces
Figure 4 for From seeing to remembering: Images with harder-to-reconstruct representations leave stronger memory traces
Viaarxiv icon

Emergent organization of receptive fields in networks of excitatory and inhibitory neurons

May 26, 2022
Viaarxiv icon

Convergence and Alignment of Gradient Descent with Random Back Propagation Weights

Jun 14, 2021
Figure 1 for Convergence and Alignment of Gradient Descent with Random Back Propagation Weights
Figure 2 for Convergence and Alignment of Gradient Descent with Random Back Propagation Weights
Figure 3 for Convergence and Alignment of Gradient Descent with Random Back Propagation Weights
Viaarxiv icon

The huge Package for High-dimensional Undirected Graph Estimation in R

Jun 26, 2020
Figure 1 for The huge Package for High-dimensional Undirected Graph Estimation in R
Figure 2 for The huge Package for High-dimensional Undirected Graph Estimation in R
Figure 3 for The huge Package for High-dimensional Undirected Graph Estimation in R
Figure 4 for The huge Package for High-dimensional Undirected Graph Estimation in R
Viaarxiv icon

Model Repair: Robust Recovery of Over-Parameterized Statistical Models

May 20, 2020
Figure 1 for Model Repair: Robust Recovery of Over-Parameterized Statistical Models
Figure 2 for Model Repair: Robust Recovery of Over-Parameterized Statistical Models
Figure 3 for Model Repair: Robust Recovery of Over-Parameterized Statistical Models
Figure 4 for Model Repair: Robust Recovery of Over-Parameterized Statistical Models
Viaarxiv icon