Unverified50% confidenceFactExact time
ZAP's training pipeline has two steps: first, bootstrapping world understanding and reasoning capability with a pretrained model, then teaching the model to move the robot on simple simulation tasks through supervised fine-tuning and GRPO.
1
Sources
50%
Confidence
Long-term
Relevance
10/27/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedZAP's core claim — that simulation training plus pretrained understanding is sufficient to support real robot control — currently comes from only a single source (rss), the 96.7% success rate was achieved on a specific task set and under a specific tool framework, and the original text does not elaborate on whether the task complexity and environmental diversity are sufficient to represent real deployment scenarios.75% similarUnverifiedZAP can complete compositional generalization tasks that do not appear in the training data, including placing test tubes into a test tube rack and selecting hole positions, handing over test tubes between two arms in mid-air, drawing X on a tic-tac-toe board, and playing chess.71% similarVerifiedA pre-trained model refers to a model that has been trained in advance on large-scale datasets, which users can directly fine-tune or call to complete specific tasks.66% similarUnverifiedMany GUI agents are trained based on Behavior Cloning, where the model learns the action distribution from human demonstration trajectories rather than the essential goal of the task66% similarUnverifiedThe computational resource requirements for robot RL training are relatively manageable, and a personal computer with an appropriate GPU can complete policy training in a simulation environment.66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1102703API
curl https://kongchang.com/api/v1/knowledge/claims/1102703MCP
get_claim(id=1102703)