MaP-WAM tackles non-Markovian robot manipulation with memory-grounded planning
Sizhe Zhao · hf · 2026-09-11
Researchers propose MaP-WAM (Memory as Plans), a world-action model for non-Markovian robotic manipulation that decouples memory-grounded planning from plan-conditioned execution.
- Uses compact episodic segment records as the memory substrate for planning
- Executes via progress-calibrated action chunks
- Maintains fixed inference latency regardless of memory length
The approach improves performance on long-horizon, history-dependent manipulation tasks.
More from Embodied
- Hybrid LLM agent scores 83.2%, beats nine VLA policies on new SO-101 robot arm benchmark — Truman Hickok · 2026-10-28
- cixliv: robot fighting may be the first realized market where bots beat average humans — cixliv · 2026-10-28
- Can the Hiwonder MentorPi T1 run fully offline as a robot pet? Pi 5 questioned — oulu2006 · 2026-10-28
- Testing Mint Modeler to Build the Menlo Research Robot — freelerobot · 2026-10-28
- Khosla-backed startup builds a 'mind-reading' cap with 100,000 tiny chips to write AI prompts — Distinct-Question-16 · 2026-10-28
- Racer turns Ray-Ban Meta glasses into an AI track coach with Opus 5.5, gains 5.3 seconds per lap — Vjeux · 2026-10-28