Unverified50% confidenceBenchmarkExact time
In the MiniGrid DoorKey-8x8 environment, a PPO agent with a reset action successfully solved the puzzle only 2 times out of 80 seeds, whereas without the reset action the success rate was 71/80.
1
Sources
50%
Confidence
Long-term
Relevance
11/3/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedIn the MiniGrid KeyCorridorS3R1 environment, the PPO agent with a reset action succeeded 0/20 times, whereas without the reset action it succeeded 20/20 times.72% similarUnverifiedIn the DoorKey environment, with advantage normalization enabled, using task reward alone solved 11 out of 20 seeds, whereas adding a tamperable satisfaction signal solved only 1.70% similarUnverifiedIn the DoorKey environment, imposing a cost of 0.1 on each reset operation restored task solving across all seeds.58% similarUnverifiedWhen a long-running Agent is split into n independent time windows, each with a success rate of p, the probability of all windows succeeding is p to the power of n, which conforms to the series system model in reliability engineering.55% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1129189API
curl https://kongchang.com/api/v1/knowledge/claims/1129189MCP
get_claim(id=1129189)