Mahyar Najibi

Computer Science · Physical Sciences

14h-index876citations52works0.62yr avg
OpenAlex↗Google Scholar↗

Accepting Students?

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

Research Topics

Advanced Neural Network Applications(21), Advanced Image and Video Retrieval Techniques(15), Domain Adaptation and Few-Shot Learning(14), Adversarial Robustness in Machine Learning(6), Robotics and Sensor-Based Localization(5)

Publications51 total

From Dense to Dynamic: Token-Difficulty Driven MoEfication of Pre-Trained LLMs
arXiv (Cornell University)·2025Open Access
QuantSpec: Self-Speculative Decoding with Hierarchical Quantized KV Cache
arXiv (Cornell University)·2025Open Access
M2R2: Mixture of Multi-Rate Residuals for Efficient Transformer Inference
arXiv (Cornell University)·2025Open Access
Speculative Streaming: Efficient and Scalable Speculative Decoding with Multi-Stream Attention
2025Open Access
Efficient Vision-Language Models by Summarizing Visual Tokens into Compact Registers
arXiv (Cornell University)·2024Open Access
CtrlSynth: Controllable Image Text Synthesis for Data-Efficient Multimodal Learning
arXiv (Cornell University)·2024Open Access
SeedLM: Compressing LLM Weights into Seeds of Pseudo-Random Generators
arXiv (Cornell University)·2024Open Access
LazyLLM: Dynamic Token Pruning for Efficient Long Context LLM Inference
arXiv (Cornell University)·2024· 1 citedOpen Access
CatLIP: CLIP-level Visual Recognition Accuracy with 2.7x Faster Pre-training on Web-scale Image-Text Data
arXiv (Cornell University)·2024Open Access
OpenELM: An Efficient Language Model Family with Open Training and Inference Framework
arXiv (Cornell University)·2024· 2 citedOpen Access
Superposition Prompting: Improving and Accelerating Retrieval-Augmented Generation
arXiv (Cornell University)·2024· 1 citedOpen Access
Speculative Streaming: Fast LLM Inference without Auxiliary Models
arXiv (Cornell University)·2024Open Access
Unsupervised 3D Perception with 2D Vision-Language Distillation for Autonomous Driving
2023· 22 cited
Unsupervised 3D Perception with 2D Vision-Language Distillation for Autonomous Driving
arXiv (Cornell University)·2023Open Access
3D Human Keypoints Estimation From Point Clouds in the Wild Without Human Labels
arXiv (Cornell University)·2023· 2 citedOpen Access
3D Human Keypoints Estimation from Point Clouds in the Wild without Human Labels
2023· 16 cited
GINA-3D: Learning to Generate Implicit Neural Assets in the Wild
2023· 16 cited
GINA-3D: Learning to Generate Implicit Neural Assets in the Wild
arXiv (Cornell University)·2023· 2 citedOpen Access
Improving the Intra-class Long-tail in 3D Detection via Rare Example Mining
arXiv (Cornell University)·2022Open Access
Motion Inspired Unsupervised Perception and Prediction in Autonomous Driving
arXiv (Cornell University)·2022· 2 citedOpen Access
page 1 of 3Next →

Frequent Co-authors

Larry S. Davis(19), Dragomir Anguelov(12), Charles R. Qi(10), Bharat Singh(8), Mohammad Rastegari(8), Yin Zhou(7), Sachin Mehta(7), Jingwei Ji(6), Xinchen Yan(6), Yin Zhou(5), Maxwell Horton(5), Ali Shafahi(4), Tom Goldstein(4), Rama Chellappa(4), Scott Ettinger(4), Boyang Deng(4), Qichen Fu(4), John P. Dickerson(3), Qingqing Cao(3), Christoph Studer(2)