Picture for Chi-Heng Lin

Chi-Heng Lin

DynaMo: Accelerating Language Model Inference with Dynamic Multi-Token Sampling

Add code
May 01, 2024
Viaarxiv icon

Balanced Data, Imbalanced Spectra: Unveiling Class Disparities with Spectral Imbalance

Add code
Feb 18, 2024
Viaarxiv icon

Half-Hop: A graph upsampling approach for slowing down message passing

Add code
Aug 17, 2023
Figure 1 for Half-Hop: A graph upsampling approach for slowing down message passing
Figure 2 for Half-Hop: A graph upsampling approach for slowing down message passing
Figure 3 for Half-Hop: A graph upsampling approach for slowing down message passing
Figure 4 for Half-Hop: A graph upsampling approach for slowing down message passing
Viaarxiv icon

The good, the bad and the ugly sides of data augmentation: An implicit spectral regularization perspective

Add code
Oct 10, 2022
Figure 1 for The good, the bad and the ugly sides of data augmentation: An implicit spectral regularization perspective
Figure 2 for The good, the bad and the ugly sides of data augmentation: An implicit spectral regularization perspective
Figure 3 for The good, the bad and the ugly sides of data augmentation: An implicit spectral regularization perspective
Figure 4 for The good, the bad and the ugly sides of data augmentation: An implicit spectral regularization perspective
Viaarxiv icon

Provable Acceleration of Heavy Ball beyond Quadratics for a Class of Polyak-Łojasiewicz Functions when the Non-Convexity is Averaged-Out

Add code
Jun 22, 2022
Figure 1 for Provable Acceleration of Heavy Ball beyond Quadratics for a Class of Polyak-Łojasiewicz Functions when the Non-Convexity is Averaged-Out
Figure 2 for Provable Acceleration of Heavy Ball beyond Quadratics for a Class of Polyak-Łojasiewicz Functions when the Non-Convexity is Averaged-Out
Figure 3 for Provable Acceleration of Heavy Ball beyond Quadratics for a Class of Polyak-Łojasiewicz Functions when the Non-Convexity is Averaged-Out
Viaarxiv icon

Drop, Swap, and Generate: A Self-Supervised Approach for Generating Neural Activity

Add code
Nov 03, 2021
Viaarxiv icon

Escaping Saddle Points Faster with Stochastic Momentum

Jun 05, 2021
Figure 1 for Escaping Saddle Points Faster with Stochastic Momentum
Figure 2 for Escaping Saddle Points Faster with Stochastic Momentum
Figure 3 for Escaping Saddle Points Faster with Stochastic Momentum
Figure 4 for Escaping Saddle Points Faster with Stochastic Momentum
Viaarxiv icon

Mine Your Own vieW: Self-Supervised Learning Through Across-Sample Prediction

Add code
Feb 19, 2021
Figure 1 for Mine Your Own vieW: Self-Supervised Learning Through Across-Sample Prediction
Figure 2 for Mine Your Own vieW: Self-Supervised Learning Through Across-Sample Prediction
Figure 3 for Mine Your Own vieW: Self-Supervised Learning Through Across-Sample Prediction
Figure 4 for Mine Your Own vieW: Self-Supervised Learning Through Across-Sample Prediction
Viaarxiv icon

Making transport more robust and interpretable by moving data through a small number of anchor points

Add code
Dec 21, 2020
Figure 1 for Making transport more robust and interpretable by moving data through a small number of anchor points
Figure 2 for Making transport more robust and interpretable by moving data through a small number of anchor points
Figure 3 for Making transport more robust and interpretable by moving data through a small number of anchor points
Figure 4 for Making transport more robust and interpretable by moving data through a small number of anchor points
Viaarxiv icon

Bayesian optimization for modular black-box systems with switching costs

Jun 04, 2020
Figure 1 for Bayesian optimization for modular black-box systems with switching costs
Figure 2 for Bayesian optimization for modular black-box systems with switching costs
Viaarxiv icon