GY
Gilad Yehudai
cs.LGstat.MLcs.NEcs.CRcs.CLcs.AIcs.CVmath.OC
On Valency
published · living versionsW_7u7j69c3·v1 · currentpublished
Improving the Expressive Power of Hybrid Sequence Models via Multi-Hop Retrieval
with Yam Eitan, Haggai Maron
1 version
Preprints & journals
31 papers in the corpus · 2019–2026Improving the Expressive Power of Hybrid Sequence Models via Multi-Hop Retrieval2610.25250v1 · Yam Eitan, Gilad Yehudai, Haggai Maron2026 · 0 citationsarXivon Valency
Lost in Tokenization: Fundamental Trade-offs in Graph Tokenization for Transformers2605.22471v2 · Maya Bechler-Speicher, Gilad Yehudai, Gil Harari et al.2026 · 0 citationsarXiv
Geometric Factual Recall in Transformers2605.12426v1 · Shauli Ravfogel, Gilad Yehudai, Joan Bruna et al.2026 · 0 citationsarXiv
Emergence of Linear Truth Encodings in Language Models2510.15804v1 · Shauli Ravfogel, Gilad Yehudai, Tal Linzen et al.2025 · 0 citationsarXiv
Provable Unlearning with Gradient Ascent on Two-Layer ReLU Neural Networks2510.14844v1 · Odelia Melamed, Gilad Yehudai, Gal Vardi2025 · 0 citationsarXiv
Querying Kernel Methods Suffices for Reconstructing their Training Data2505.19019v1 · Daniel Barzilai, Yuval Margalit, Eitan Gronich et al.2025 · 0 citationsarXiv
Compositional Reasoning with Transformers, RNNs, and Chain of Thought2503.01544v2 · Gilad Yehudai, Noah Amsel, Joan Bruna2025 · 0 citationsarXiv
Depth-Width tradeoffs in Algorithmic Reasoning of Graph Tasks with Transformers2503.01805v3 · Gilad Yehudai, Clayton Sanford, Maya Bechler-Speicher et al.2025 · 0 citationsarXiv
Logarithmic Width Suffices for Robust Memorization2502.11162v1 · Amitsour Egosi, Gilad Yehudai, Ohad Shamir2025 · 0 citationsarXiv
On the Reconstruction of Training Data from Group Invariant Networks2411.16458v1 · Ran Elbaz, Gilad Yehudai, Meirav Galun et al.2024 · 0 citationsarXiv
On the Benefits of Rank in Attention Layers2407.16153v1 · Noah Amsel, Gilad Yehudai, Joan Bruna2024 · 0 citationsarXiv
Reconstructing Training Data From Real World Models Trained with Transfer Learning2407.15845v1 · Yakir Oz, Gilad Yehudai, Gal Vardi et al.2024 · 0 citationsarXiv
When Can Transformers Count to n?2407.15160v3 · Gilad Yehudai, Haim Kaplan, Guy Dar et al.2024 · 1 citationarXiv
MALT Powers Up Adversarial Attacks2407.02240v1 · Odelia Melamed, Gilad Yehudai, Adi Shamir2024 · 0 citationsarXiv
RedEx: Beyond Fixed Representation Methods via Convex Optimization2401.07606v1 · Amit Daniely, Mariano Schain, Gilad Yehudai2024 · 0 citationsarXiv
Locally Optimal Descent for Dynamic Stepsize Scheduling2311.13877v1 · Gilad Yehudai, Alon Cohen, Amit Daniely et al.2023 · 0 citationsarXiv
Deconstructing Data Reconstruction: Multiclass, Weight Decay and General Losses2307.01827v2 · Gon Buzaglo, Niv Haim, Gilad Yehudai et al.2023 · 6 citationsarXiv
From Tempered to Benign Overfitting in ReLU Neural Networks2305.15141v3 · Guy Kornowski, Gilad Yehudai, Ohad Shamir2023 · 4 citationsarXiv
Reconstructing Training Data from Multiclass Neural Networks2305.03350v1 · Gon Buzaglo, Niv Haim, Gilad Yehudai et al.2023 · 3 citationsarXiv
Adversarial Examples Exist in Two-Layer ReLU Networks for Low Dimensional Linear Subspaces2303.00783v2 · Odelia Melamed, Gilad Yehudai, Gal Vardi2023 · 1 citationarXiv
Reconstructing Training Data from Trained Neural Networks2206.07758v3 · Niv Haim, Gal Vardi, Gilad Yehudai et al.2022 · 48 citationsarXiv
Gradient Methods Provably Converge to Non-Robust Networks2202.04347v2 · Gal Vardi, Gilad Yehudai, Ohad Shamir2022 · 6 citationsarXiv
Width is Less Important than Depth in ReLU Neural Networks2202.03841v2 · Gal Vardi, Gilad Yehudai, Ohad Shamir2022 · 6 citationsarXiv
On the Optimal Memorization Power of ReLU Neural Networks2110.03187v1 · Gal Vardi, Gilad Yehudai, Ohad Shamir2021 · 3 citationsarXiv
Learning a Single Neuron with Bias Using Gradient Descent2106.01101v2 · Gal Vardi, Gilad Yehudai, Ohad Shamir2021 · 2 citationsarXiv
The Connection Between Approximation, Depth Separation and Learnability in Neural Networks2102.00434v2 · Eran Malach, Gilad Yehudai, Shai Shalev-Shwartz et al.2021 · 5 citationsarXiv
From Local Structures to Size Generalization in Graph Neural Networks2010.08853v3 · Gilad Yehudai, Ethan Fetaya, Eli Meirom et al.2020 · 25 citationsarXiv
The Effects of Mild Over-parameterization on the Optimization Landscape of Shallow ReLU Neural Networks2006.01005v2 · Itay Safran, Gilad Yehudai, Ohad Shamir2020 · 5 citationsarXiv
Proving the Lottery Ticket Hypothesis: Pruning is All You Need2002.00585v1 · Eran Malach, Gilad Yehudai, Shai Shalev-Shwartz et al.2020 · 127 citationsarXiv
Learning a Single Neuron with Gradient Methods2001.05205v3 · Gilad Yehudai, Ohad Shamir2020 · 12 citationsarXiv
On the Power and Limitations of Random Features for Understanding Neural Networks1904.00687v4 · Gilad Yehudai, Ohad Shamir2019 · 47 citationsarXiv
Career total: 34 works. 31 are in this corpus.
Profile built from the corpus for this byline.
Author records are still filling in while the Hub is in alpha. If this is your page, you'll be able to claim it soon. Spot a mistake? Tell us.