Samet Oymak
Associate Professor, EECS – Electrical and Computer Engineering
Studies principled methods and theory for machine learning, including the foundations of transformers and resource-efficient, trustworthy learning.
Samet Oymak's lab at Michigan works on learning theory & optimization, trustworthy & safe ML and efficient training & inference. 38 papers since 2023, 30 at top venues, mostly NeurIPS and AAAI. Applications close .
At a glance
- papers, 2023–2026
- 10/ year
- at T1 venues
- 79%
- PhD students
- —
- MS students
- —
- alumni listed
- —
Papers by yearfrom DBLP, 2023–2026
- 2 papers2026 · T1 2 · collecting
- 13 papers2025 · T1 11
- 13 papers2024 · T1 10
- 10 papers2023 · T1 7
T1 venuesOtherVenues: NeurIPS, AAAI, ICML, AISTATS, ICLR
Research focuspaper titles that use each topic's words
- Learning theory & optimization14 papers
- Efficient training & inference3 papers
- Trustworthy & safe ML2 papers
- principled1 paper
- resource-efficient0 papers
Paper list
From DBLP.
20262 papers
202513 papers
- CPALA Case Study of Low Ranked Self-Expressive Structures in Neural Network Representations (opens in a new tab)
- CVPR (T1)AdMiT: Adaptive Multi-Source Tuning in Dynamic Environments (opens in a new tab)
- NeurIPS (T1)Attention with Trained Embeddings Provably Selects Important Tokens (opens in a new tab)
- NeurIPS (T1)BREAD: Branched Rollouts from Expert Anchors Bridge SFT & RL for Reasoning (opens in a new tab)
- ICML (T1)Everything Everywhere All at Once: LLMs can In-Context Learn Multiple Tasks in Superposition (opens in a new tab)
- NeurIPS (T1)Extrapolation by Association: Length Generalization Transfer In Transformers (opens in a new tab)
- MILCOMFLASH: Federated Learning Across Simultaneous Heterogeneities (opens in a new tab)
- ICLR (T1)High-dimensional Analysis of Knowledge Distillation: Weak-to-Strong Generalization and Scaling Laws (opens in a new tab)
- AAAI (T1)On the Power of Convolution-Augmented Transformer (opens in a new tab)
- AISTATS (T1)Provable Benefits of Task-Specific Prompts for In-context Learning (opens in a new tab)
- ICML (T1)Test-Time Training Provably Improves Transformers as In-context Learners (opens in a new tab)
- AAAI (T1)TimePFN: Effective Multivariate Time Series Forecasting with Synthetic Data (opens in a new tab)
202413 papers
- AAAI (T1)A Score-Based Deterministic Diffusion Algorithm with Smooth Scores for General Distributions (opens in a new tab)
- ICML (T1)Can Mamba Learn How To Learn? A Comparative Study on In-Context Learning Tasks (opens in a new tab)
- AAAI (T1)Class-Attribute Priors: Adapting Optimization to Heterogeneity and Fairness Objective (opens in a new tab)
- NeurIPS (T1)CONTRAST: Continual Multi-source Adaptation to Dynamic Distributions (opens in a new tab)
- WACVEffective Restoration of Source Knowledge in Continual Test Time Adaptation (opens in a new tab)
- NeurIPS (T1)Efficient Contextual LLM Cascades through Budget-Constrained Policy Learning (opens in a new tab)
- NeurIPS (T1)Fine-grained Analysis of In-context Linear Estimation: Data, Architecture, and Beyond (opens in a new tab)
- ICML (T1)From Self-Attention to Markov Models: Unveiling the Dynamics of Generative Transformers (opens in a new tab)
- AISTATS (T1)Mechanics of Next Token Prediction with Self-Attention (opens in a new tab)
- CDCPrediction for Dynamical Systems via Transfer Learning (opens in a new tab)
- SIGSPATIAL/GISPyneapple-L: Scalable Expressive Learning-based Spatial Analysis (opens in a new tab)
- NeurIPS (T1)Selective Attention: Enhancing Transformer through Principled Context Control (opens in a new tab)
202310 papers
- IEEE Control. Syst. Lett.Can Transformers Learn Optimal Filtering for Unknown Systems? (opens in a new tab)
- NeurIPS (T1)Dissecting Chain-of-Thought: Compositionality through In-Context Filtering and Learning (opens in a new tab)
- L4DCLearning on Manifolds: Universal Approximations Properties using Geometric Controllability Conditions for Neural ODEs (opens in a new tab)
- NeurIPS (T1)Max-Margin Token Selection in Attention Mechanism (opens in a new tab)
- ICASSP (T1)On The Fairness of Multitask Representation Learning (opens in a new tab)
- ICML (T1)On the Role of Attention in Prompt-tuning (opens in a new tab)
- AAAI (T1)Provable Pathways: Learning Multiple Tasks over Multiple Paths (opens in a new tab)
- AAAI (T1)Stochastic Contextual Bandits with Long Horizon Rewards (opens in a new tab)
- EMS@SIGCOMMText-to-3D Generative AI on Mobile Devices: Measurements and Optimizations (opens in a new tab)
- ICML (T1)Transformers as Algorithms: Generalization and Stability in In-context Learning (opens in a new tab)
Where alumni went
Alumni are not listed on the lab homepage.
Apply to MichiganFall 2027
Michigan CSE Computer Science PhD
- Round
- Single round
- Opens
- —
- GRE
- See program page
- English
- See program page
- Fee
- See program page
Not yet confirmed for this intake.
Similar labsnearest by research summary
- CNSJTU
Wenrui Dai
- Learning theory & optimization
- CNZhejiang
Jiachi Chen
- Trustworthy & safe ML
- CHEPFL
Pascal Frossard
- Trustworthy & safe ML
- USUC Berkeley
Kannan Ramchandran
- Signal processing & information theory
- CHETH Zurich
Thomas Hofmann
- Learning theory & optimization
- USMichigan
Gabriel Poesia
- Reasoning & neuro-symbolic AI
Your notes
Save this lab to keep notes.
Data
Some numbers could not be verified:
- The member list could not be found on the lab site.
- The alumni list could not be found on the lab site.
- The lab site and the paper record do not agree.
- Collected
- Verified
- Papers
- DBLP
- Members
- Not collected
- Alumni
- Not collected
Tiers are venue classifications, not lab rankings.
Michigancloses