Picture for Kamesh Krishnamurthy

Kamesh Krishnamurthy

The Relational Bottleneck as an Inductive Bias for Efficient Abstraction

Add code
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

Trainability, Expressivity and Interpretability in Gated Neural ODEs

Add code
Jul 12, 2023
Figure 1 for Trainability, Expressivity and Interpretability in Gated Neural ODEs
Figure 2 for Trainability, Expressivity and Interpretability in Gated Neural ODEs
Figure 3 for Trainability, Expressivity and Interpretability in Gated Neural ODEs
Figure 4 for Trainability, Expressivity and Interpretability in Gated Neural ODEs
Viaarxiv icon

Theory of gating in recurrent neural networks

Add code
Aug 29, 2020
Figure 1 for Theory of gating in recurrent neural networks
Figure 2 for Theory of gating in recurrent neural networks
Figure 3 for Theory of gating in recurrent neural networks
Figure 4 for Theory of gating in recurrent neural networks
Viaarxiv icon

Gating creates slow modes and controls phase-space complexity in GRUs and LSTMs

Add code
Jan 31, 2020
Figure 1 for Gating creates slow modes and controls phase-space complexity in GRUs and LSTMs
Figure 2 for Gating creates slow modes and controls phase-space complexity in GRUs and LSTMs
Figure 3 for Gating creates slow modes and controls phase-space complexity in GRUs and LSTMs
Figure 4 for Gating creates slow modes and controls phase-space complexity in GRUs and LSTMs
Viaarxiv icon