Unverified50% confidenceOpinionExact time
In a civilization sandbox, pure ε-greedy random exploration is extremely inefficient when the action space is about 5000, and hand-written priorities can greatly accelerate early learning.
1
Sources
50%
Confidence
Long-term
Relevance
10/23/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe ε-greedy policy selects actions randomly with probability ε and selects the action with the highest current Q-value with probability 1-ε; ε is typically set to 1.0 at the beginning of training and then linearly decayed to 0.01 or 0.1 over millions of frames.67% similarVerifiedThe learning rate determines how large each step is: too large causes overshooting the valley bottom and divergence, while too small makes training extremely time-consuming.67% similarVerifiedDuring cross-domain fine-tuning, an excessively large learning rate will destroy pre-trained weights, while an excessively small one will result in almost no learning, requiring a careful trade-off.67% similarUnverifiedMuch tacit knowledge that affects results (such as the need for a special learning rate warm-up strategy in the first few hundred steps of an optimizer) is difficult to fully capture in standardized tables.67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1088928API
curl https://kongchang.com/api/v1/knowledge/claims/1088928MCP
get_claim(id=1088928)