Signals 4 · free daily AI digest

VC-Attention: Value Smoothing and Softmax Casting for Low-bit Attention

Paper recorded by Signals 4 on 2026-09-14 in cs.CV. Abstract reproduced from arXiv; link to the original below.

Published 2026-09-14 on arXiv · recorded by Signals 4 on 2026-09-15

Category: cs.CV · 计算机视觉 · first seen 2026-09-15

Abstract

Diffusion Transformers deliver state-of-the-art video generation, but their long spatiotemporal sequences make attention the dominant deployment cost, and a deployable low-bit kernel must be accurate and fast. Accuracy is limited by outliers: a block's quantization scale is set by its largest entries, leaving typical entries confined to a narrow range of representable values. Prior work smooths qu

Read on arXiv →

#48 most recent of 237 cs.CV papers we have recorded · ↑ newer: Integrating Multi-view Multi-light Surface Reconstruction into Cultura · ↓ older: SURE-Map: Self-Correcting Streaming Geometric Foundation Model
Cite this page: VC-Attention: Value Smoothing and Softmax Casting for Low-bit Attention: the #48 most recent of 237 cs.CV papers we have recorded (as of 2026-09-14). Source: Signals 4 (Signals API) — https://data.jiangzhang.ca/signals4/t/papers/vc-attention-value-smoothing-and-softmax-casting-for-low-bit-attention.html
Free to quote with attribution to “Signals 4 (Signals API)”. Machine-readable: papers.json
Related: More cs.CV papers · arXiv signals · All papers · Today in AI
Get 4 AI signals a day by email — free.
Subscribe free → See all plans →
Get 4 AI signals a day by email — free
All models · All repos · By company · Daily editions