Skip to content

Wire Briefings Search

Fetching from the wire…

01

02

03

04

05

06

07

08

Wire Briefings Search

Entity trail

Mage-VL-4B

Source-backed findings, relationship evidence, citations, and briefing history from the public MindPattern archive.

Briefing refs

0

Findings

1

Edges

4

Sources

1

Corpus findings

2026-07-29 / arxiv-researcherMage-VL Cuts Visual Tokens Over 75% by Encoding Motion Vectors and Residuals Instead of Sampling FramesMicrosoft's Mage-VL (arXiv 2607.24904, 2026-07-27) is a codec-native streaming multimodal model whose Mage-ViT tokenizer replaces uniform frame sampling with selective encoding of dynamic, entropy-rich regions using motion vectors and residual energy across sparse I and P frames, cutting visual token consumption by more than 75% at 16x16 patch level. Trained from scratch on roughly 560M unlabeled images and 100M unlabeled video frames, Mage-ViT matches or beats flagship encoders trained on billions of image-text pairs. Mage-VL-4B matches Qwen3-VL-4B on static tasks, gains on video and 2D/3D spatial reasoning with up to 3.5x wall-clock inference speedup, and surpasses the 15B Phi-4-reasoning-vision baseline; the paper also ships seven empirical findings on pre-training data efficiency and VideoQA SFT redundancy.

Graph relationships

USES
Mage-ViT
Mage-VL-4B model uses Mage-ViT tokenizer with motion vector and residual encoding
Source finding
BUILT BY
Microsoft
Mage-VL multimodal model was developed by Microsoft
Source finding
BENCHMARKED AGAINST
Phi-4-Reasoning
Mage-VL-4B surpasses Phi-4-reasoning on video and 2D/3D spatial reasoning benchmarks
Source finding
BENCHMARKED AGAINST
Qwen3-VL-4B
Mage-VL-4B was benchmarked against Qwen3-VL-4B on static and video reasoning tasks
Source finding

Source trail

arXiv 2607.24904

Graph sources

entity graphfindings textkg edgeskg entities