ViTRM
ai model→ stable
Vision Tiny Recursion Model
ViTRM, developed by researchers, is a recursive vision transformer that applies a single 3-layer block repeatedly, achieving competitive performance with dramatically reduced parameters.
1Total Mentions
+0.60Sentiment (Very Positive)
0.0%Velocity (7d)
First seen: Mar 23, 2026Last active: Mar 23, 2026
Signal Radar
Five-axis snapshot of this entity's footprint
Loading radar…
Mentions × Lab Attention
Weekly mentions (solid) and average article relevance (dotted)
mentionsrelevance
Loading timeline…
Timeline
1- Research MilestoneMar 23, 2026
Proposed Vision Tiny Recursion Model achieving competitive CIFAR performance with 84x fewer parameters than Vision Transformers.
View source- parameter reduction:
- 84x
- benchmarks:
- CIFAR-10, CIFAR-100
Relationships
No relationships mapped yet.
Recent Articles
No articles found for this entity.
Predictions
No predictions linked to this entity.
AI Discoveries
No AI agent discoveries for this entity.
Sentiment History
Positive sentiment
Negative sentiment
Range: -1 to +1
| Week | Avg Sentiment | Mentions |
|---|---|---|
| 2026-W13 | 0.60 | 1 |