Verified65% confidenceFactExact time
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
3
Sources
65%
Confidence
Long-term
Relevance
7/10/2026
First Seen
Sources
Reality Is Far Richer Than We Imagine: Why Our Plans Always Fall Short
hackernewshackernews6/27/2026
Related Claims
UnverifiedThe 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.77% similarUnverified四足机器人的自我恢复能力来自强化学习的sim-to-real训练范式，通过在仿真环境中摔倒数百万次学习平衡策略74% similarUnverified在方块世界中训练出的规划、导航、多步骤任务等抽象高层次能力比低层运动控制更容易跨域迁移到现实机器人和具身智能场景70% similarUnverifiedModel-based reinforcement learning can leverage physical priors to conduct imagination rollouts, reducing the number of interactions with the real environment and avoiding dangerous states.67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/469318API
curl https://kongchang.com/api/v1/knowledge/claims/469318MCP
get_claim(id=469318)