Matthew L. Leavitt

According to our database, Matthew L. Leavitt authored at least 10 papers between 2020 and 2024.

Collaborative distances:

• Dijkstra number of four.
• Erdős number of four.

2024

Dynamic Masking Rate Schedules for MLM Pretraining.

Zachary Ankner, Naomi Saphra, Davis W. Blalock, Jonathan Frankle, Matthew L. Leavitt

Proceedings of the 18th Conference of the European Chapter of the Association for Computational Linguistics, 2024

2023

Sudden Drops in the Loss: Syntax Acquisition, Phase Transitions, and Simplicity Bias in MLMs.

Angelica Chen, Ravid Schwartz-Ziv, Kyunghyun Cho, Matthew L. Leavitt, Naomi Saphra

CoRR, 2023

On the special role of class-selective neurons in early training.

Omkar Ranadive, Nikhil Thakurdesai, Ari S. Morcos, Matthew L. Leavitt, Stéphane Deny

CoRR, 2023

Knowledge Distillation for Efficient Sequences of Training Runs.

Xingyu Liu, Alex Leonardi, Lu Yu, Chris Gilmer-Hill, Matthew L. Leavitt, Jonathan Frankle

CoRR, 2023

2022

Reduce, Reuse, Recycle: Improving Training Efficiency with Distillation.

Cody Blakeney, Jessica Zosa Forde, Jonathan Frankle, Ziliang Zong, Matthew L. Leavitt

CoRR, 2022

2021

ConViT: Improving Vision Transformers with Soft Convolutional Inductive Biases.

Stéphane d'Ascoli, Hugo Touvron, Matthew L. Leavitt, Ari S. Morcos, Giulio Biroli, Levent Sagun

Proceedings of the 38th International Conference on Machine Learning, 2021

Selectivity considered harmful: evaluating the causal impact of class selectivity in DNNs.

Matthew L. Leavitt, Ari S. Morcos

Proceedings of the 9th International Conference on Learning Representations, 2021

2020

Towards falsifiable interpretability research.

Matthew L. Leavitt, Ari S. Morcos

CoRR, 2020

Linking average- and worst-case perturbation robustness via class selectivity and dimensionality.

Matthew L. Leavitt, Ari S. Morcos

CoRR, 2020

On the relationship between class selectivity, dimensionality, and robustness.

Matthew L. Leavitt, Ari S. Morcos

CoRR, 2020