Llama 3.2 VISION Tested - Shockingly Censored! 🤬

Llama 3.2 Vision performs well on basic image descriptions.

The model successfully described simple images, demonstrating its capability in basic visual recognition tasks.

Llama 3.2 Vision shows significant censorship in its responses.

The model fails to identify people in images and solve simple captchas, indicating a high level of censorship compared to previous models.

Llama 3.2 Vision's performance is inconsistent.

While it succeeded in some tasks, its failures in others highlight inconsistency in performance.

Complex tasks reveal Llama 3.2's limitations.

The model struggled with more complex tasks, such as identifying Waldo in a detailed image.