Yiqiao Zhong
Scholar

Yiqiao Zhong

Google Scholar ID: in8fQ10AAAAJ
Assistant Professor, University of Wisconsin--Madison
Interpretability of LLMsDeep Learning TheoryMachine LearningStatistics
Homepage↗Google Scholar↗
Citations & Impact
All-time
Citations
1,350
 
H-index
13
 
i10-index
15
 
Publications
20
 
Co-authors
9
list available
Contact
No contact links provided.
Publications
7 items
Task Vector Geometry Underlies Dual Modes of Task Inference in Transformers
2026
Cited
0
How Does Unfaithful Reasoning Emerge from Autoregressive Training? A Study of Synthetic Experiments
2026
Cited
0
Shattered Compositionality: Counterintuitive Learning Dynamics of Transformers for Arithmetic
2026
Cited
0
Unifying Attention Heads and Task Vectors via Hidden State Geometry in In-Context Learning
2025
Cited
0
A statistical theory of overfitting for imbalanced classification
2025
Cited
0
Assessing and improving reliability of neighbor embedding methods: a map-continuity perspective
arXiv.org · 2024
Cited
0
Out-of-distribution generalization via composition: a lens through induction heads in Transformers
arXiv.org · 2024
Cited
2
Resume (English only)
Co-authors
9 total
Jianqing Fan
Jianqing Fan
Professor of Statistics, Professor of Finance, Princeton University
Kaizheng Wang
Kaizheng Wang
Columbia University
Co-author 3
Co-author 3
Cong Ma
Cong Ma
University of Chicago
Andrea Montanari
Andrea Montanari
John D. and Sigrid Banks Professor, Statistics and Mathematics, Stanford University
Nicolas Boumal
Nicolas Boumal
Ecole Polytechnique Fédérale de Lausanne (EPFL)
Liwei Wang
Liwei Wang
Professor, Peking University
Chi Jin
Chi Jin
Assistant Professor, Princeton University