YL
Yin Tat Lee
cs.DSmath.OCcs.LGstat.MLcs.CLcs.AIcs.CRcs.NAmath.NAmath.PR
On Valency
published · living versionsW_qrshuaqb·v1 · currentpublished
Early science acceleration experiments with GPT-5
with Sébastien Bubeck, Christian Coester, Ronen Eldan, Timothy Gowers +9
1 version
Preprints & journals
100 papers in the corpus · 2013–2025OpenAI GPT-5 System Card2601.03267v2 · Aaditya Singh, Adam Fry, Adam Perelman et al.2025 · 26 citationsarXiv
Early science acceleration experiments with GPT-52511.16072v1 · S'ebastien Bubeck, Christian Coester, Ronen Eldan et al.2025 · 3 citationsarXivon Valency
Pseudo-Asynchronous Local SGD: Robust and Efficient Data-Parallel Training2504.18454v2 · Hiroki Naganuma, Xinzhi Zhang, Man-Chung Yue et al.2025 · 0 citationsarXiv
Phi-4 Technical Report2412.08905v1 · Marah Abdin, Jyoti Aneja, Harkirat Behl et al.2024 · 31 citationsarXiv
Phi-3 Technical Report: A Highly Capable Language Model Locally on Your Phone2404.14219v4 · Marah Abdin, Jyoti Aneja, Hany Awadalla et al.2024 · 165 citationsarXiv
Improving the Bit Complexity of Communication for Distributed Convex Optimization2403.19146v1 · Mehrdad Ghadiri, Yin Tat Lee, Swati Padmanabhan et al.2024 · 1 citationarXiv
Differentially Private Synthetic Data via Foundation Model APIs 2: Text2403.01749v2 · Chulin Xie, Zinan Lin, Arturs Backurs et al.2024 · 2 citationsarXiv
Can Generalist Foundation Models Outcompete Special-Purpose Tuning? Case Study in Medicine2311.16452v1 · Harsha Nori, Yin Tat Lee, Sheng Zhang et al.2023 · 178 citationsarXiv
Positional Description Matters for Transformers Arithmetic2311.14737v1 · Ruoqi Shen, S'ebastien Bubeck, Ronen Eldan et al.2023 · 3 citationsarXiv
Textbooks Are All You Need II: phi-1.5 technical report2309.05463v1 · Yuanzhi Li, S'ebastien Bubeck, Ronen Eldan et al.2023 · 49 citationsarXiv
Textbooks Are All You Need2306.11644v2 · Suriya Gunasekar, Yi Zhang, Jyoti Aneja et al.2023 · 138 citationsarXiv
MathChat: Converse to Tackle Challenging Math Problems with LLM Agents2306.01337v3 · Yiran Wu, Feiran Jia, Shaokun Zhang et al.2023 · 15 citationsarXiv
Automatic Prompt Optimization with "Gradient Descent" and Beam Search2305.03495v2 · Reid Pryzant, Dan Iter, Jerry Li et al.2023 · 191 citationsarXiv
Convex Minimization with Integer Minima in Time2304.03426v2 · Haotian Jiang, Yin Tat Lee, Zhao Song et al.2023 · 0 citationsarXiv
Sparks of Artificial General Intelligence: Early experiments with GPT-42303.12712v5 · S'ebastien Bubeck, Varun Chandrasekaran, Ronen Eldan et al.2023 · 1,523 citationsarXiv
NN-Adapter: Efficient Domain Adaptation for Black-Box Language Models2302.10879v1 · Yangsibo Huang, Daogao Liu, Zexuan Zhong et al.2023 · 4 citationsarXiv
Algorithmic Aspects of the Log-Laplace Transform and a Non-Euclidean Proximal Sampler2302.06085v3 · Sivakanth Gopi, Yin Tat Lee, Daogao Liu et al.2023 · 1 citationarXiv
ReSQueing Parallel and Private Stochastic Convex Optimization2301.00457v2 · Yair Carmon, Arun Jambulapati, Yujia Jin et al.2023 · 0 citationsarXiv
Learning threshold neurons via the "edge of stability"2212.07469v2 · Kwangjun Ahn, S'ebastien Bubeck, Sinho Chewi et al.2022 · 0 citationsarXiv
Exploring the Limits of Differentially Private Deep Learning with Group-wise Clipping2212.01539v1 · Jiyan He, Xuechen Li, Da Yu et al.2022 · 6 citationsarXiv
Upper and Lower Bounds on the Smoothed Complexity of the Simplex Method2211.11860v3 · Sophie Huiberts, Yin Tat Lee, Xinzhi Zhang2022 · 2 citationsTheoretiCS, Volume 4 (October 15, 2025) theoretics:13604
Condition-number-independent convergence rate of Riemannian Hamiltonian Monte Carlo with numerical integrators2210.07219v2 · Yunbum Kook, Yin Tat Lee, Ruoqi Shen et al.2022 · 3 citationsarXiv
A Slightly Improved Bound for the KLS Constant2208.11644v2 · Arun Jambulapati, Yin Tat Lee, Santosh S. Vempala2022 · 4 citationsarXiv
Decomposable Non-Smooth Convex Optimization with Nearly-Linear Gradient Oracle Complexity2208.03811v1 · Sally Dong, Haotian Jiang, Yin Tat Lee et al.2022 · 1 citationarXiv
Private Convex Optimization in General Norms2207.08347v2 · Sivakanth Gopi, Yin Tat Lee, Daogao Liu et al.2022 · 4 citationsarXiv
Nested Dissection Meets IPMs: Planar Min-Cost Flow in Nearly-Linear Time2205.01562v1 · Sally Dong, Yu Gao, Gramoz Goranci et al.2022 · 3 citationsarXiv
Private Convex Optimization via Exponential Mechanism2203.00263v2 · Sivakanth Gopi, Yin Tat Lee, Daogao Liu2022 · 12 citationsarXiv
Short-step Methods Are Not Strongly Polynomial-Time2201.02768v1 · Manru Zong, Yin Tat Lee, Man-Chung Yue2022 · 3 citationsarXiv
A gradient sampling method with complexity guarantees for Lipschitz functions in high and low dimensions2112.06969v2 · Damek Davis, Dmitriy Drusvyatskiy, Yin Tat Lee et al.2021 · 4 citationsarXiv
Faster Maxflow via Improved Dynamic Spectral Vertex Sparsifiers2112.00722v1 · Jan van den Brand, Yu Gao, Arun Jambulapati et al.2021 · 18 citationsarXiv
Computing Lewis Weights to High Precision2110.15563v1 · Maryam Fazel, Yin Tat Lee, Swati Padmanabhan et al.2021 · 5 citationsarXiv
Differentially Private Fine-tuning of Language Models2110.06500v2 · Da Yu, Saurabh Naik, Arturs Backurs et al.2021 · 101 citationsarXiv
Tutorial on the Robust Interior Point Method2108.04734v1 · Yin Tat Lee, Santosh S. Vempala2021 · 4 citationsarXiv
Mechanistic model-driven exometabolomic characterisation of human dopaminergic neuronal metabolism10.1101/2021.06.30.450562v2 · Preciat, G., Wegrzyn, A., Moreno, E. L. et al.2021 · 11 citationsbioRxiv
Lower Bounds on Metropolized Sampling Methods for Well-Conditioned Distributions2106.05480v2 · Yin Tat Lee, Ruoqi Shen, Kevin Tian2021 · 8 citationsarXiv
Numerical Composition of Differential Privacy2106.02848v3 · Sivakanth Gopi, Yin Tat Lee, Lukas Wutschitz2021 · 58 citationsarXiv
The Power of Sampling: Dimension-free Risk Bounds in Private ERM2105.13637v4 · Yin Tat Lee, Daogao Liu, Zhou Lu2021 · 1 citationarXiv
Private Non-smooth Empirical Risk Minimization and Stochastic Convex Optimization in Subquadratic Steps2103.15352v2 · Janardhan Kulkarni, Yin Tat Lee, Daogao Liu2021 · 6 citationsarXiv
Fast and Memory Efficient Differentially Private-SGD via JL Projections2102.03013v1 · Zhiqi Bu, Sivakanth Gopi, Janardhan Kulkarni et al.2021 · 9 citationsarXiv
Minimum Cost Flows, MDPs, and -Regression in Nearly Linear Time for Dense Instances2101.05719v2 · Jan van den Brand, Yin Tat Lee, Yang P. Liu et al.2021 · 13 citationsarXiv
A Nearly-Linear Time Algorithm for Linear Programs with Small Treewidth: A Multiscale Representation of Robust Central Path2011.05365v3 · Sally Dong, Yin Tat Lee, Guanghao Ye2020 · 17 citationsarXiv
Structured Logconcave Sampling with a Restricted Gaussian Oracle2010.03106v4 · Yin Tat Lee, Ruoqi Shen, Kevin Tian2020 · 9 citationsarXiv
A Faster Interior Point Method for Semidefinite Programming2009.10217v1 · Haotian Jiang, Tarun Kathuria, Yin Tat Lee et al.2020 · 63 citationsarXiv
Reducing Isotropy and Volume to KLS: Faster Rounding and Volume Algorithms2008.02146v3 · He Jia, Aditi Laddha, Yin Tat Lee et al.2020 · 2 citationsarXiv
Composite Logconcave Sampling with a Restricted Gaussian Oracle2006.05976v1 · Ruoqi Shen, Kevin Tian, Yin Tat Lee2020 · 3 citationsarXiv
Network size and weights size for memorization with two-layers neural networks2006.02855v2 · S'ebastien Bubeck, Ronen Eldan, Yin Tat Lee et al.2020 · 16 citationsarXiv
An Improved Cutting Plane Method for Convex Optimization, Convex-Concave Games and its Applications2004.04250v1 · Haotian Jiang, Yin Tat Lee, Zhao Song et al.2020 · 52 citationsarXiv
Acceleration with a Ball Optimization Oracle2003.08078v1 · Yair Carmon, Arun Jambulapati, Qijia Jiang et al.2020 · 7 citationsarXiv
Positive Semidefinite Programming: Mixed, Parallel, and Width-Independent2002.04830v3 · Arun Jambulapati, Yin Tat Lee, Jerry Li et al.2020 · 10 citationsarXiv
Logsmooth Gradient Concentration and Tighter Runtimes for Metropolized Hamiltonian Monte Carlo2002.04121v3 · Yin Tat Lee, Ruoqi Shen, Kevin Tian2020 · 13 citationsarXiv
Career total: 155 works. 100 are in this corpus.Showing the 50 most recent.
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.