Ziwei Ji

39 papers A* 13B 1C 1Journal 22Unranked 2
YearRankTypeTitle / Venue / Authors
2026 J jnl
DynaWeb: Model-Based Reinforcement Learning of Web Agents.
CoRR
Hang Ding, Peidong Liu, Junqiao Wang, Ziwei Ji, Meng Cao, Rongzhao Zhang, Lynn Ai, Eric Yang, Tianyu Shi, Lei Yu
2025 J jnl
Mixture-of-Recursions: Learning Dynamic Recursive Depths for Adaptive Token-Level Computation.
CoRR
Sangmin Bae, Yujin Kim, Reza Bayat, Sungnyun Kim, Jiyoun Ha, Tal Schuster, Adam Fisch, Hrayr Harutyunyan, Ziwei Ji, Aaron C. Courville, Se-Young Yun
2025 A* conf
Relaxed Recursive Transformers: Effective Parameter Sharing with Layer-wise LoRA.
ICLR
Sangmin Bae, Adam Fisch, Hrayr Harutyunyan, Ziwei Ji, Seungyeon Kim, Tal Schuster
2024 J jnl
Efficient Document Ranking with Learnable Late Interactions.
CoRR
Ziwei Ji, Himanshu Jain, Andreas Veit, Sashank J. Reddi, Sadeep Jayasumana, Ankit Singh Rawat, Aditya Krishna Menon, Felix X. Yu, Sanjiv Kumar
2024 J jnl
Relaxed Recursive Transformers: Effective Parameter Sharing with Layer-wise LoRA.
CoRR
Sangmin Bae, Adam Fisch, Hrayr Harutyunyan, Ziwei Ji, Seungyeon Kim, Tal Schuster
2024 A* conf
Think before you speak: Training Language Models With Pause Tokens.
ICLR
Sachin Goyal, Ziwei Ji, Ankit Singh Rawat, Aditya Krishna Menon, Sanjiv Kumar, Vaishnavh Nagarajan
2023 J jnl
Depth Dependence of μP Learning Rates in ReLU MLPs.
CoRR
Samy Jelassi, Boris Hanin, Ziwei Ji, Sashank J. Reddi, Srinadh Bhojanapalli, Sanjiv Kumar
2023 J jnl
Think before you speak: Training Language Models With Pause Tokens.
CoRR
Sachin Goyal, Ziwei Ji, Ankit Singh Rawat, Aditya Krishna Menon, Sanjiv Kumar, Vaishnavh Nagarajan
2022 A* conf
Actor-critic is implicitly biased towards high entropy optimal policies.
ICLR
Yuzheng Hu, Ziwei Ji, Matus Telgarsky
2022 A* conf
Agnostic Learnability of Halfspaces via Logistic Loss.
ICML
Ziwei Ji, Kwangjun Ahn, Pranjal Awasthi, Satyen Kale, Stefani Karp
2022 J jnl
Agnostic Learnability of Halfspaces via Logistic Loss.
CoRR
Ziwei Ji, Kwangjun Ahn, Pranjal Awasthi, Satyen Kale, Stefani Karp
2022 J jnl
Convex Analysis at Infinity: An Introduction to Astral Space.
CoRR
Miroslav Dudík, Ziwei Ji, Robert E. Schapire, Matus Telgarsky
2022 A* conf
Reproducibility in Optimization: Theoretical Framework and Limits.
NeurIPS
Kwangjun Ahn, Prateek Jain, Ziwei Ji, Satyen Kale, Praneeth Netrapalli, Gil I. Shamir
2022 J jnl
Reproducibility in Optimization: Theoretical Framework and Limits.
CoRR
Kwangjun Ahn, Prateek Jain, Ziwei Ji, Satyen Kale, Praneeth Netrapalli, Gil I. Shamir
2021 J jnl
Actor-critic is implicitly biased towards high entropy optimal policies.
CoRR
Yuzheng Hu, Ziwei Ji, Matus Telgarsky
2021 B conf
Characterizing the implicit bias via a primal-dual analysis.
ALT
Ziwei Ji, Matus Telgarsky
2021 A* conf
Early-stopped neural networks are consistent.
NeurIPS
Ziwei Ji, Justin D. Li, Matus Telgarsky
2021 J jnl
Early-stopped neural networks are consistent.
CoRR
Ziwei Ji, Justin D. Li, Matus Telgarsky
2021 J jnl
Fast Margin Maximization via Dual Acceleration.
CoRR
Ziwei Ji, Nathan Srebro, Matus Telgarsky
2021 A* conf
Fast margin maximization via dual acceleration.
ICML
Ziwei Ji, Nathan Srebro, Matus Telgarsky
2021 A* conf
Generalization bounds via distillation.
ICLR
Daniel Hsu, Ziwei Ji, Matus Telgarsky, Lan Wang
2021 J jnl
Generalization bounds via distillation.
CoRR
Daniel Hsu, Ziwei Ji, Matus Telgarsky, Lan Wang
2020 A* conf
Directional convergence and alignment in deep learning.
NeurIPS
Ziwei Ji, Matus Telgarsky
2020 J jnl
Directional convergence and alignment in deep learning.
CoRR
Ziwei Ji, Matus Telgarsky
2020 A* conf
Gradient descent follows the regularization path for general losses.
COLT
Ziwei Ji, Miroslav Dudík, Robert E. Schapire, Matus Telgarsky
2020 J jnl
Gradient descent follows the regularization path for general losses.
CoRR
Ziwei Ji, Miroslav Dudík, Robert E. Schapire, Matus Telgarsky
2020 A* conf
Neural tangent kernels, transportation mappings, and universal approximation.
ICLR
Ziwei Ji, Matus Telgarsky, Ruicheng Xian
2020 A* conf
Polylogarithmic width suffices for gradient descent to achieve arbitrarily small test error with shallow ReLU networks.
ICLR
Ziwei Ji, Matus Telgarsky
2019 J jnl
A refined primal-dual analysis of the implicit bias.
CoRR
Ziwei Ji, Matus Telgarsky
2019 — conf
Gradient descent aligns the layers of deep linear networks.
ICLR (Poster)
Ziwei Ji, Matus Telgarsky
2019 J jnl
Neural tangent kernels, transportation mappings, and universal approximation.
CoRR
Ziwei Ji, Matus Telgarsky, Ruicheng Xian
2019 J jnl
Polylogarithmic width suffices for gradient descent to achieve arbitrarily small test error with shallow ReLU networks.
CoRR
Ziwei Ji, Matus Telgarsky
2019 A* conf
The implicit bias of gradient descent on nonseparable data.
COLT
Ziwei Ji, Matus Telgarsky
2018 J jnl
Gradient descent aligns the layers of deep linear networks.
CoRR
Ziwei Ji, Matus Telgarsky
2018 J jnl
Risk and parameter convergence of logistic regression.
CoRR
Ziwei Ji, Matus Telgarsky
2018 — conf
Social Welfare and Profit Maximization from Revealed Preferences.
WINE
Ziwei Ji, Ruta Mehta, Matus Telgarsky
2017 J jnl
Social Welfare and Profit Maximization from Revealed Preferences.
CoRR
Ziwei Ji, Ruta Mehta, Matus Telgarsky
2017 J jnl
Wikidata Vandalism Detection - The Loganberry Vandalism Detector at WSDM Cup 2017.
CoRR
Qi Zhu, Hongwei Ng, Liyuan Liu, Ziwei Ji, Bingjie Jiang, Jiaming Shen, Huan Gui
2016 C conf
The Beachcombers' Problem: Walking and Searching from an Inner Point of a Line.
LATA
Yu Chen, Xiaotie Deng, Ziwei Ji, Chao Liao