Verified70% confidenceFactExact time
The sim-to-real gap refers to the phenomenon where policies trained in simulation often suffer significant performance degradation when transferred to the real world, and domain randomization and system identification are commonly used mitigation techniques.
4
Sources
70%
Confidence
Long-term
Relevance
8/21/2026
First Seen
Sources
Related Claims
Unverified域随机化（Domain Randomization）和系统辨识等技术在2016-2020年出现，大幅缩小了仿真与现实的差距，DeepMind、OpenAI等机构成功将仿真训练策略迁移到真实硬件78% similarVerified增强学习领域用'sim-to-real gap'描述在模拟环境训练的机器人策略因现实物理差异而在真实世界失效的现象77% similarUnverified训练-部署鸿沟指模型因训练与部署环境的数据分布偏移、软件依赖版本差异或硬件指令集不同而导致的性能下降73% similarUnverifiedThe training and validation of physical AI relies heavily on the Sim-to-Real Transfer closed loop.71% similarUnverifiedWhen models trained on game-rendered images are directly applied to real-world scenes, performance often degrades significantly due to domain differences.70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/782730API
curl https://kongchang.com/api/v1/knowledge/claims/782730MCP
get_claim(id=782730)