Team RETINA
Using a generalized Sylvester equation and Gemini 3 Pro preview, we're able to bridge the gap between videos and LLMs.
Core trick
Uses a generalized Sylvester equation to map video representations into an LLM input space for Gemini 3 Pro.
Description
Using a generalized Sylvester equation and Gemini 3 Pro preview, we're able to bridge the gap between videos and LLMs.
The linked repo returns 404, so it was deleted or made private.