Back-of-envelope: 1M context can hold 16+ hours of per-minute visual snapshots
YouJiacheng · x · 2026-11-23
YouJiacheng does the math: an image tokenizes to 256 tokens, and with modern LLMs' 1M context, allocating 256k tokens to visual snapshots yields 1000 minutes—over 16 hours of visual memory, far larger than any existing VLM's memory. This counters the claim that visual tokens are too expensive to scale context for robot history.
Related event: VLA Memory Design: LLM for Long-Term History, Sliding Window for Control(3 posts)→
More from Embodied
- Municipal Pipe Robot R2D2 Rescues Kitten Trapped in Utah Storm Drain — AndrewCritchPhD · 2026-11-23
- Robot's incredible footwork takes champion title, Air Street Capital founder reacts — nathanbenaich · 2026-11-23
- ModRetro Chromatic: Codex Edition handheld ships with a programmable cartridge — Yamapama · 2026-11-23
- Dad builds a working 3D-printed Pokedex with ESP32 and local YOLO11n, all via Hermes — Teknium · 2026-11-23
- Tesla Optimus plays piano as humanoid dexterity hits a new milestone — dioscuri · 2026-11-23
- Stable JEPA could auto-chunk time and only invoke the predictor on surprises — MarcJSchmidt · 2026-11-23