YouJiacheng: 16-hour visual snapshot capacity dwarfs any existing VLM memory

YouJiacheng · x · 2026-11-23

YouJiacheng reiterates that the per-minute visual snapshot scheme (16+ hours of memory) is already far larger than any existing VLM's memory capacity, further countering concerns about visual token cost.

Related event: VLA Memory Design: LLM for Long-Term History, Sliding Window for Control(3 posts)→

Original post →

More from Embodied

Embodied channel →