Uni-SafeBench

product→ stable

Uni-SafeBench, developed by Scale AI and Universal Robots, is a standardized safety evaluation framework for AI systems that reveals multimodal models have 30-50% higher failure rates than specialized counterparts.

1Total Mentions

-0.30Sentiment (Negative)

+1.2%Velocity (7d)

First seen: Apr 3, 2026Last active: 12h ago

Timeline

Research MilestoneApr 1, 2026
Study showing unified multimodal models have 30-50% higher safety failure rates
View source

Relationships

Uses

→
GPT-4V
ai model1 source80% conf.
→
Gemini
product1 source80% conf.

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