Byte Latent Transformer: Patches Scale Better Than Tokens
Paper • 2412.09871 • Published • 108
Architectural Proposals
-
-
Causal Diffusion Transformers for Generative Modeling
Paper • 2412.12095 • Published • 23 -
Tensor Product Attention Is All You Need
Paper • 2501.06425 • Published • 90 -
TransMLA: Multi-head Latent Attention Is All You Need
Paper • 2502.07864 • Published • 69 -
Transformers without Normalization
Paper • 2503.10622 • Published • 172 -
LSNet: See Large, Focus Small
Paper • 2503.23135 • Published • 11 -
DDT: Decoupled Diffusion Transformer
Paper • 2504.05741 • Published • 77 -
Latent Diffusion Autoencoders: Toward Efficient and Meaningful Unsupervised Representation Learning in Medical Imaging
Paper • 2504.08635 • Published • 4 -
D^2iT: Dynamic Diffusion Transformer for Accurate Image Generation
Paper • 2504.09454 • Published • 11 -
Efficient Generative Model Training via Embedded Representation Warmup
Paper • 2504.10188 • Published • 12 -
Softpick: No Attention Sink, No Massive Activations with Rectified Softmax
Paper • 2504.20966 • Published • 31 -
Group Downsampling with Equivariant Anti-aliasing
Paper • 2504.17258 • Published • 9 -
Latent Flow Transformer
Paper • 2505.14513 • Published • 29 -
LaTtE-Flow: Layerwise Timestep-Expert Flow-based Transformer
Paper • 2506.06952 • Published • 9 -
Marrying Autoregressive Transformer and Diffusion with Multi-Reference Autoregression
Paper • 2506.09482 • Published • 45 -
From Bytes to Ideas: Language Modeling with Autoregressive U-Nets
Paper • 2506.14761 • Published • 17 -
Energy-Based Transformers are Scalable Learners and Thinkers
Paper • 2507.02092 • Published • 69 -
Dynamic Chunking for End-to-End Hierarchical Sequence Modeling
Paper • 2507.07955 • Published • 27 -
Region-based Cluster Discrimination for Visual Representation Learning
Paper • 2507.20025 • Published • 20 -
PixNerd: Pixel Neural Field Diffusion
Paper • 2507.23268 • Published • 52 -
Local Scale Equivariance with Latent Deep Equilibrium Canonicalizer
Paper • 2508.14187 • Published • 4 -
Artificial Hippocampus Networks for Efficient Long-Context Modeling
Paper • 2510.07318 • Published • 32 -
Virtual Width Networks
Paper • 2511.11238 • Published • 39 -
mHC: Manifold-Constrained Hyper-Connections
Paper • 2512.24880 • Published • 322 -
Scaling Embeddings Outperforms Scaling Experts in Language Models
Paper • 2601.21204 • Published • 102 -
Scaling Text-to-Image Diffusion Transformers with Representation Autoencoders
Paper • 2601.16208 • Published • 55 -
Nested Learning: The Illusion of Deep Learning Architectures
Paper • 2512.24695 • Published • 45 -
Locality-Attending Vision Transformer
Paper • 2603.04892 • Published • 7 -
Scale Space Diffusion
Paper • 2603.08709 • Published • 16 -
V-Co: A Closer Look at Visual Representation Alignment via Co-Denoising
Paper • 2603.16792 • Published • 3 -
ViT-AdaLA: Adapting Vision Transformers with Linear Attention
Paper • 2603.16063 • Published • 2