{"type":"rich","version":"1.0","provider_name":"Transistor","provider_url":"https://transistor.fm","author_name":"Pretrained","title":"DeepSeek's New LLM Architectures","html":"<iframe width=\"100%\" height=\"180\" frameborder=\"no\" scrolling=\"no\" seamless src=\"https://share.transistor.fm/e/4269f5bd\"></iframe>","width":"100%","height":180,"duration":3608,"description":"Pierce and Richard break down DeepSeek's latest model architecture moves in Manifold-Constrained Hyper Connections and Engram memory. Are these conceptually sound? Will they hop the pond over to US frontier labs?","thumbnail_url":"https://img.transistorcdn.com/8veBHYJ1tFjtWlv9ET3YGaLijqZK5MYE6tVoUbwgKaw/rs:fill:0:0:1/w:400/h:400/q:60/mb:500000/aHR0cHM6Ly9pbWct/dXBsb2FkLXByb2R1/Y3Rpb24udHJhbnNp/c3Rvci5mbS8yYTRi/YTQ5Zjk4ZjIzMmU2/YzRiMWZjN2E5ZmJk/NzNjNi5wbmc.webp","thumbnail_width":300,"thumbnail_height":300}