English
Related papers

Related papers: Transformers without Normalization

200 papers

We introduce a deep and light-weight transformer, DeLighT, that delivers similar or better performance than standard transformer-based models with significantly fewer parameters. DeLighT more efficiently allocates parameters both (1) within…

Machine Learning · Computer Science 2021-02-15 Sachin Mehta , Marjan Ghazvininejad , Srinivasan Iyer , Luke Zettlemoyer , Hannaneh Hajishirzi

The success of deep learning is inseparable from normalization layers. Researchers have proposed various normalization functions, and each of them has both advantages and disadvantages. In response, efforts have been made to design a…

Machine Learning · Computer Science 2024-02-20 Zikai Zhou , Shuo Zhang , Ziruo Wang , Huanran Chen

Neural population activity is theorized to reflect an underlying dynamical structure. This structure can be accurately captured using state space models with explicit dynamics, such as those based on recurrent neural networks (RNNs).…

Neurons and Cognition · Quantitative Biology 2023-07-21 Joel Ye , Chethan Pandarinath

The successful training of deep neural networks requires addressing challenges such as overfitting, numerical instabilities leading to divergence, and increasing variance in the residual stream. A common solution is to apply regularization…

Machine Learning · Computer Science 2025-11-20 Jörg K. H. Franke , Urs Spiegelhalter , Marianna Nezhurina , Jenia Jitsev , Frank Hutter , Michael Hefenbrock

Foundation models achieve state-of-the-art performance across different tasks, but their size and computational demands raise concerns about accessibility and sustainability. Existing efficiency methods often require additional retraining…

Normalization techniques have only recently begun to be exploited in supervised learning tasks. Batch normalization exploits mini-batch statistics to normalize the activations. This was shown to speed up training and result in better…

Machine Learning · Computer Science 2017-03-08 Mengye Ren , Renjie Liao , Raquel Urtasun , Fabian H. Sinz , Richard S. Zemel

We present a parameter-efficient Diffusion Transformer (DiT) for generating 200bp cell-type-specific regulatory DNA sequences. By replacing the U-Net backbone of DNA-Diffusion with a transformer denoiser equipped with a 2D CNN input…

Machine Learning · Computer Science 2026-03-12 Jonathan Liu , Kia Ghods

Diffusion models have achieved remarkable success across a range of generative tasks. Recent efforts to enhance diffusion model architectures have reimagined them as a form of multi-task learning, where each task corresponds to a denoising…

Computer Vision and Pattern Recognition · Computer Science 2024-07-11 Byeongjun Park , Hyojun Go , Jin-Young Kim , Sangmin Woo , Seokil Ham , Changick Kim

Recent work studying the generalization of diffusion models with UNet-based denoisers reveals inductive biases that can be expressed via geometry-adaptive harmonic bases. However, in practice, more recent denoising networks are often based…

Computer Vision and Pattern Recognition · Computer Science 2024-10-29 Jie An , De Wang , Pengsheng Guo , Jiebo Luo , Alexander Schwing

It is a common belief that training deep transformers from scratch requires large datasets. Consequently, for small datasets, people usually use shallow and simple additional layers on top of pre-trained models during fine-tuning. This work…

Computation and Language · Computer Science 2021-06-01 Peng Xu , Dhruv Kumar , Wei Yang , Wenjie Zi , Keyi Tang , Chenyang Huang , Jackie Chi Kit Cheung , Simon J. D. Prince , Yanshuai Cao

Transformer becomes prevalent in computer vision, especially for high-level vision tasks. However, adopting Transformer in the generative adversarial network (GAN) framework is still an open yet challenging problem. In this paper, we…

Computer Vision and Pattern Recognition · Computer Science 2023-06-14 Rui Xu , Xiangyu Xu , Kai Chen , Bolei Zhou , Chen Change Loy

Recent developments in large-scale pre-trained text-to-image diffusion models have significantly improved the generation of high-fidelity images, particularly with the emergence of diffusion transformer models (DiTs). Among diffusion…

Computer Vision and Pattern Recognition · Computer Science 2025-04-08 Xudong Lu , Aojun Zhou , Ziyi Lin , Qi Liu , Yuhui Xu , Renrui Zhang , Xue Yang , Junchi Yan , Peng Gao , Hongsheng Li

We introduce GrounDiT, a novel training-free spatial grounding technique for text-to-image generation using Diffusion Transformers (DiT). Spatial grounding with bounding boxes has gained attention for its simplicity and versatility,…

Computer Vision and Pattern Recognition · Computer Science 2024-11-04 Phillip Y. Lee , Taehoon Yoon , Minhyuk Sung

Transformer is a state-of-the-art model in the field of natural language processing (NLP). Current NLP models primarily increase the number of transformers to improve processing performance. However, this technique requires a lot of…

Computation and Language · Computer Science 2023-10-18 Woohyeon Moon , Taeyoung Kim , Bumgeun Park , Dongsoo Har

Recently, the tokens of images share the same static data flow in many dense networks. However, challenges arise from the variance among the objects in images, such as large variations in the spatial scale and difficulties of recognition…

Computer Vision and Pattern Recognition · Computer Science 2023-08-14 Yuchen Ma , Zhengcong Fei , Junshi Huang

The Transformer is widely used in natural language processing tasks. To train a Transformer however, one usually needs a carefully designed learning rate warm-up stage, which is shown to be crucial to the final performance but will slow…

Machine Learning · Computer Science 2020-06-30 Ruibin Xiong , Yunchang Yang , Di He , Kai Zheng , Shuxin Zheng , Chen Xing , Huishuai Zhang , Yanyan Lan , Liwei Wang , Tie-Yan Liu

Diffusion Transformers (DiTs) have achieved state-of-the-art performance in image and video generation, but their success comes at the cost of heavy computation. This inefficiency is largely due to the fixed tokenization process, which uses…

Computer Vision and Pattern Recognition · Computer Science 2026-02-20 Dahye Kim , Deepti Ghadiyaram , Raghudeep Gadde

With sequentially stacked self-attention, (optional) encoder-decoder attention, and feed-forward layers, Transformer achieves big success in natural language processing (NLP), and many variants have been proposed. Currently, almost all…

Computation and Language · Computer Science 2021-03-08 Jinhua Zhu , Lijun Wu , Yingce Xia , Shufang Xie , Tao Qin , Wengang Zhou , Houqiang Li , Tie-Yan Liu

Many neural network architectures are known to be Turing Complete, and can thus, in principle implement arbitrary algorithms. However, Transformers are unique in that they can implement gradient-based learning algorithms under simple…

Machine Learning · Computer Science 2024-06-05 Xiang Cheng , Yuxin Chen , Suvrit Sra

Batch normalization is a key component of most image classification models, but it has many undesirable properties stemming from its dependence on the batch size and interactions between examples. Although recent work has succeeded in…

Computer Vision and Pattern Recognition · Computer Science 2021-02-12 Andrew Brock , Soham De , Samuel L. Smith , Karen Simonyan