These experiments led to two key discoveries, according to the paper. Tuning only the self-attention projection layers (SA Proj), the part of the model that helps it decide which input elements to focus on, allowed the models to learn new tasks with little or no measurable forgetting. Also, what initially appeared as forgotten knowledge often resurfaced when the model was later trained on another specialized task.
The way we see it, the real race for AI video hasn't begun. Our new identity, Mirage, reflects our expanded vision and commitment to redefining the video category, starting with short-form video, through frontier AI research and models, CEO Gaurav Misra told TechCrunch.
We delved into the five pretraining datasets of 34 multimodal vision-language models, analyzing the distribution and composition of concepts within, generating over 300GB of data artifacts that we publicly release.