paper-with-me

Papers

Flash3D: Super-scaling Point Transformers through Joint Hardware-Geometry Locality

2024-12-21 · CVPR 2025 1 · Liyan Chen, Gregory P. Meyer, Zaiwei Zhang, Eric M. Wolff, Paul Vernaza

Recent efforts recognize the power of scale in 3D learning (e.g. PTv3) and attention mechanisms (e.g. FlashAttention). However, current point cloud backbones fail to holistically unify geometric locality, attention mechanisms, and GPU architectures in one view. In this paper, we introduce Flash3D Transformer, which aligns geometric locality and GPU tiling through a principled locality mechanism based on Perfect Spatial Hashing (PSH). The common alignment with GPU tiling naturally fuses our PSH locality mechanism with FlashAttention at negligible extra cost. This mechanism affords flexible design choices throughout the backbone that result in superior downstream task results. Flash3D outperforms state-of-the-art PTv3 results on benchmark datasets, delivering a 2.25x speed increase and 2.4x memory efficiency boost. This efficiency enables scaling to wider attention scopes and larger models without additional overhead. Such scaling allows Flash3D to achieve even higher task accuracies than PTv3 under the same compute budget.

📄 PDF Abstract BibTeX arXiv:2412.16481

Code (1)

liyanc/flash3dtransformer 공식 구현 pytorch

Tasks

GPU

Methods 이 논문이 사용한 방법론

Attention 설명 없음
Linear Layer A Linear Layer is a projection $\mathbf{XW + b}$.
BPE Byte Pair Encoding, or BPE, is a subword segmentation algorithm that encodes rare and unknown words as sequences of subword units. The intuition is that various word…
Absolute Position Encodings Absolute Position Encodings are a type of position embeddings for [Transformer-based models] where positional encodings are…
Dense Connections Dense Connections, or Fully Connected Connections, are a type of layer in a deep neural network that use a linear operation where every input is connected to every output…
Multi-Head Attention 설명 없음
SPEED The monocular depth estimation (MDE) is the task of estimating depth from a single frame. This information is an essential knowledge in many computer vision tasks such as scene…
Position-Wise Feed-Forward Layer 설명 없음

Similar Papers 제목 키워드 기반

Rank-Factorized Implicit Neural Bias: Scaling Super-Resolution Transformer with FlashAttention

2026-03-06 · Dongheon Lee, Seokju Yun, Jaegyun Im, Youngmin Ro arxiv

Recent Super-Resolution~(SR) methods mainly adopt Transformers for their strong long-range modeling capability and exceptional representational capacity. However, most SR Transformers rely heavily on relative positional …

Long-range modeling

Hungry Hungry Hippos: Towards Language Modeling with State Space Models

2022-12-28 · Daniel Y. Fu, Tri Dao, Khaled K. Saab, Armin W. Thomas 외

State space models (SSMs) have demonstrated state-of-the-art sequence modeling performance in some modalities, but underperform attention in language modeling. Moreover, despite scaling nearly linearly in sequence length…

8kCoreference ResolutionFew-Shot LearningLanguage Modeling+6

Flash Multi-Head Feed-Forward Network

2025-12-07 · Minshen Zhang, Xiang Hu, Jianguo Li, Wei Wu 외 arxiv

We explore Multi-Head FFN (MH-FFN) as a replacement of FFN in the Transformer architecture, motivated by the structural similarity between single-head attention and FFN. While multi-head mechanisms enhance expressivity i…

Tiled Flash Linear Attention: More Efficient Linear RNN and xLSTM Kernels

2025-03-18 · Maximilian Beck, Korbinian Pöppel, Phillip Lippe, Sepp Hochreiter

Linear RNNs with gating recently demonstrated competitive performance compared to Transformers in language modeling. Although their linear compute scaling in sequence length offers theoretical runtime advantages over Tra…

GPULanguage ModelingLanguage ModellingMamba

Dispatch-Aware Ragged Attention for Pruned Vision Transformers

2026-04-16 · Seifeldin Abdellatif, Ahmad Almasri arxiv

Token pruning methods for Vision Transformers (ViTs) promise quadratic reductions in attention FLOPs by dropping uninformative patches. Yet standard variable-length attention APIs -- including FlashAttention-2's varlen a…