Utku Evci

Computer Science · Physical Sciences

11h-index1.1kcitations33works0.52yr avg
OpenAlex↗Google Scholar↗

Accepting Students?

No reports yet. Know if this professor is taking students?

Research Topics

Domain Adaptation and Few-Shot Learning(16), Advanced Neural Network Applications(14), Machine Learning and Data Classification(9), Stochastic Gradient Optimization Techniques(6), Multimodal Machine Learning Applications(6)

Publications33 total

Spark Transformer: Reactivating Sparsity in FFN and Attention
arXiv (Cornell University)·2025Open Access
Compression Scaling Laws:Unifying Sparsity and Quantization
arXiv (Cornell University)·2025Open Access
The Journey Matters: Average Parameter Count over Pre-training Unifies Sparse and Dense Scaling Laws
arXiv (Cornell University)·2025Open Access
Learning Parameter Sharing with Tensor Decompositions and Sparsity
arXiv (Cornell University)·2024· 1 citedOpen Access
Towards Optimal Adapter Placement for Efficient Transfer Learning
arXiv (Cornell University)·2024Open Access
Progressive Gradient Flow for Robust N:M Sparsity Training in Transformers
arXiv (Cornell University)·2024· 2 citedOpen Access
Scaling Laws for Sparsely-Connected Foundation Models
arXiv (Cornell University)·2023Open Access
Dynamic Sparse Training with Structured Sparsity
arXiv (Cornell University)·2023· 4 citedOpen Access
JaxPruner: A concise library for sparsity research
arXiv (Cornell University)·2023· 1 citedOpen Access
The Dormant Neuron Phenomenon in Deep Reinforcement Learning
arXiv (Cornell University)·2023· 7 citedOpen Access
Scaling Vision Transformers to 22 Billion Parameters
arXiv (Cornell University)·2023· 117 citedOpen Access
Training Recipe for N:M Structured Sparsity with Decaying Pruning Mask
arXiv (Cornell University)·2022· 3 citedOpen Access
Gradient Flow in Sparse Neural Networks and How Lottery Tickets Win
Proceedings of the AAAI Conference on Artificial Intelligence·2022· 36 citedOpen Access
The State of Sparse Training in Deep Reinforcement Learning
arXiv (Cornell University)·2022· 5 citedOpen Access
GradMax: Growing Neural Networks using Gradient Information
arXiv (Cornell University)·2022· 9 citedOpen Access
Head2Toe: Utilizing Intermediate Representations for Better Transfer Learning
arXiv (Cornell University)·2022· 16 citedOpen Access
A Unified Few-Shot Classification Benchmark to Compare Transfer and Meta Learning Approaches
Neural Information Processing Systems·2021· 3 cited
Practical Real Time Recurrent Learning with a Sparse Approximation
International Conference on Learning Representations·2021· 3 cited
Comparing Transfer and Meta Learning Approaches on a Unified Few-Shot\n Classification Benchmark
arXiv (Cornell University)·2021· 11 citedOpen Access
Comparing Transfer and Meta Learning Approaches on a Unified Few-Shot Classification Benchmark
arXiv (Cornell University)·2021· 8 citedOpen Access
page 1 of 2Next →

Frequent Co-authors

Vincent Dumoulin(6), Hugo Larochelle(6), Yann Dauphin(6), Erich Elsen(6), Neil Houlsby(6), Ross Goroshin(5), Pablo Samuel Castro(5), Suvinay Subramanian(5), Jacob Menick(4), Xiaohua Zhai(4), Yani Ioannou(4), Amir Yazdanbakhsh(4), Léon Bottou(3), Trevor Gale(3), Sylvain Gelly(3), Sheng-Chun Kao(3), Shivani Agrawal(3), Elias Frantar(3), Dan Alistarh(3), Eleni Triantafillou(2)