Yuvion VL: A Multimodal Foundation Model for Adversarial Content and AI Safety

Name: Yuvion VL: A Multimodal Foundation Model for Adversarial Content and AI Safety
Uploaded: 2026-10-22T11:36:53.083542+00:00
Duration: 1062 s
Description: Yuvion VL is a family of multimodal large language models specifically designed for content and AI safety, treating safety as an adversarial problem. It uses an automated data pipeline and a three-stage training process, including Confuse-then-Contrast Fine-Tuning (C2FT), to make the model robust against harmful content while maintaining general capabilities.

Yuvion VL is a family of multimodal large language models specifically designed for content and AI safety, treating safety as an adversarial problem. It uses an automated data pipeline and a three-stage training process, including Confuse-then-Contrast Fine-Tuning (C2FT), to make the model robust against harmful content while maintaining general capabilities.

Open the full episode