Eran Malach

Publication Activity (10 Years)

Years Active: 2017-2024
Publications (10 Years): 44

Top Topics

Linearly Separable

Provably Correct

Top Venues

Publications

Avital Shafran, Eran Malach, Thomas Ristenpart, Gil Segev, Stefano Tessaro
Is ML-Based Cryptanalysis Inherently Limited? Simulating Cryptographic Adversaries via Gradient-Based Methods. IACR Cryptol. ePrint Arch. 2024 (2024)
Kaiying Hou, David Brandfonbrener, Sham M. Kakade, Samy Jelassi, Eran Malach
Universal Length Generalization with Turing Programs. CoRR (2024)
Benjamin L. Edelman, Ezra Edelman, Surbhi Goel, Eran Malach, Nikolaos Tsilivis
The Evolution of Statistical Induction Heads: In-Context Learning Markov Chains. CoRR (2024)
Avital Shafran, Eran Malach, Thomas Ristenpart, Gil Segev, Stefano Tessaro
Is ML-Based Cryptanalysis Inherently Limited? Simulating Cryptographic Adversaries via Gradient-Based Methods. CRYPTO (6) (2024)
Depen Morwani, Itai Shapira, Nikhil Vyas, Eran Malach, Sham M. Kakade, Lucas Janson
A New Perspective on Shampoo's Preconditioner. CoRR (2024)
Edwin Zhang, Vincent Zhu, Naomi Saphra, Anat Kleiman, Benjamin L. Edelman, Milind Tambe, Sham M. Kakade, Eran Malach
Transcendence: Generative Models Can Outperform The Experts That Train Them. CoRR (2024)
Samy Jelassi, David Brandfonbrener, Sham M. Kakade, Eran Malach
Repeat After Me: Transformers are Better than State Space Models at Copying. CoRR (2024)
Eran Malach
Auto-Regressive Next-Token Predictors are Universal Learners. CoRR (2023)
Benjamin L. Edelman, Surbhi Goel, Sham M. Kakade, Eran Malach, Cyril Zhang
Pareto Frontiers in Deep Feature Learning: Data, Compute, Width, and Luck. NeurIPS (2023)
Benjamin L. Edelman, Surbhi Goel, Sham M. Kakade, Eran Malach, Cyril Zhang
Pareto Frontiers in Neural Feature Learning: Data, Compute, Width, and Luck. CoRR (2023)
Gal Kaplun, Andrey Gurevich, Tal Swisa, Mazor David, Shai Shalev-Shwartz, Eran Malach
SubTuning: Efficient Finetuning for Multi-Task Learning. CoRR (2023)
Etay Livne, Gal Kaplun, Eran Malach, Shai Shalev-Schwatz
Corgi^2: A Hybrid Offline-Online Approach To Storage-Aware Data Shuffling For SGD. CoRR (2023)
Gal Kaplun, Eran Malach, Preetum Nakkiran, Shai Shalev-Shwartz
Knowledge Distillation: Bad Models Can Be Good Role Models. CoRR (2022)
Alon Brutzkus, Amir Globerson, Eran Malach, Alon Regev Netser, Shai Shalev-Shwartz
Efficient Learning of CNNs using Patch Based Features. ICML (2022)
Boaz Barak, Benjamin L. Edelman, Surbhi Goel, Sham M. Kakade, Eran Malach, Cyril Zhang
Hidden Progress in Deep Learning: SGD Learns Parities Near the Computational Limit. CoRR (2022)
Boaz Barak, Benjamin L. Edelman, Surbhi Goel, Sham M. Kakade, Eran Malach, Cyril Zhang
Hidden Progress in Deep Learning: SGD Learns Parities Near the Computational Limit. NeurIPS (2022)
Gal Kaplun, Eran Malach, Preetum Nakkiran, Shai Shalev-Shwartz
Knowledge Distillation: Bad Models Can Be Good Role Models. NeurIPS (2022)
Eran Malach, Shai Shalev-Shwartz
When Hardness of Approximation Meets Hardness of Learning. J. Mach. Learn. Res. 23 (2022)
Eran Malach, Gilad Yehudai, Shai Shalev-Shwartz, Ohad Shamir
The Connection Between Approximation, Depth Separation and Learnability in Neural Networks. COLT (2021)
Eran Malach, Pritish Kamath, Emmanuel Abbe, Nathan Srebro
Quantifying the Benefit of Using Differentiable Learning over Tangent Kernels. CoRR (2021)
Emmanuel Abbe, Pritish Kamath, Eran Malach, Colin Sandon, Nathan Srebro
On the Power of Differentiable Learning versus PAC and SQ Learning. CoRR (2021)
Eran Malach, Gilad Yehudai, Shai Shalev-Shwartz, Ohad Shamir
The Connection Between Approximation, Depth Separation and Learnability in Neural Networks. CoRR (2021)
Emmanuel Abbe, Pritish Kamath, Eran Malach, Colin Sandon, Nathan Srebro
On the Power of Differentiable Learning versus PAC and SQ Learning. NeurIPS (2021)
Eran Malach, Shai Shalev-Shwartz
Computational Separation Between Convolutional and Fully-Connected Networks. ICLR (2021)
Eran Malach, Pritish Kamath, Emmanuel Abbe, Nathan Srebro
Quantifying the Benefit of Using Differentiable Learning over Tangent Kernels. ICML (2021)
Eran Malach, Shai Shalev-Shwartz
When Hardness of Approximation Meets Hardness of Learning. CoRR (2020)
Eran Malach, Gilad Yehudai, Shai Shalev-Shwartz, Ohad Shamir
Proving the Lottery Ticket Hypothesis: Pruning is All You Need. CoRR (2020)
Amit Daniely, Eran Malach
Learning Parities with Neural Networks. NeurIPS (2020)
Eran Malach, Shai Shalev-Shwartz
The Implications of Local Correlation on Learning Some Deep Functions. NeurIPS (2020)
Amit Daniely, Eran Malach
Learning Parities with Neural Networks. CoRR (2020)
Eran Malach, Gilad Yehudai, Shai Shalev-Shwartz, Ohad Shamir
Proving the Lottery Ticket Hypothesis: Pruning is All You Need. ICML (2020)
Eran Malach, Shai Shalev-Shwartz
Computational Separation Between Convolutional and Fully-Connected Networks. CoRR (2020)
Alon Brutzkus, Amit Daniely, Eran Malach
ID3 Learns Juntas for Smoothed Product Distributions. COLT (2020)
Eran Malach, Shai Shalev-Shwartz
Is Deeper Better only when Shallow is Good? CoRR (2019)
Eran Malach, Shai Shalev-Shwartz
Is Deeper Better only when Shallow is Good? NeurIPS (2019)
Jonathan Fiat, Eran Malach, Shai Shalev-Shwartz
Decoupling Gating from Linearity. CoRR (2019)
Alon Brutzkus, Amit Daniely, Eran Malach
ID3 Learns Juntas for Smoothed Product Distributions. CoRR (2019)
Alon Brutzkus, Amit Daniely, Eran Malach
On the Optimality of Trees Generated by ID3. CoRR (2019)
Eran Malach, Shai Shalev-Shwartz
Learning Boolean Circuits with Neural Networks. CoRR (2019)
Eran Malach, Shai Shalev-Shwartz
A Provably Correct Algorithm for Deep Learning that Actually Works. CoRR (2018)
Alon Brutzkus, Amir Globerson, Eran Malach, Shai Shalev-Shwartz
SGD Learns Over-parameterized Networks that Provably Generalize on Linearly Separable Data. ICLR (Poster) (2018)
Eran Malach, Shai Shalev-Shwartz
Decoupling "when to update" from "how to update". CoRR (2017)
Eran Malach, Shai Shalev-Shwartz
Decoupling "when to update" from "how to update". NIPS (2017)
Alon Brutzkus, Amir Globerson, Eran Malach, Shai Shalev-Shwartz
SGD Learns Over-parameterized Networks that Provably Generalize on Linearly Separable Data. CoRR (2017)