Verified75% confidenceFactExact time
Reward shaping refers to adding extra, denser guiding reward signals on top of the original sparse reward to help agents learn faster and more correctly.
3
Sources
75%
Confidence
Long-term
Relevance
9/27/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedReward engineering is essentially a highly iterative process: design the reward, observe behavior, discover loopholes, correct the reward, and repeat.75% similarUnverifiedPotential-based reward shaping guarantees that the optimal policy remains unchanged.73% similarUnverifiedThe NGU algorithm superimposes external rewards with two types of intrinsic rewards: a short-term reward based on intra-episode novelty and a long-term reward based on lifetime novelty.72% similarUnverifiedWithout the potential function shaping constraint, artificially set shaping rewards easily induce the policy to pursue shaping bypasses rather than truly completing the task (reward hacking).72% similarUnverifiedCheckpoint partial rewards densify sparse signals by setting milestone-style rewards at intermediate nodes in long tasks, helping gradient propagation.71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/953202API
curl https://kongchang.com/api/v1/knowledge/claims/953202MCP
get_claim(id=953202)