Unverified50% confidenceFactExact time
The probability distribution returned by the Choice type is usually softmax-normalized and sums to 1.
1
Sources
50%
Confidence
Long-term
Relevance
11/17/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedConfidence values typically originate from the model's softmax probability distribution over each option — a large probability gap yields a high confidence value, while a uniform distribution yields a low one.75% similarUnverifiedSoftmax converts the raw scores (logits) of the network's last layer into a probability distribution, where each class's output value is between 0 and 1 and sums to 1.68% similarUnverifiedWhen T=1 the output is a standard softmax distribution; when T>1 the probability distribution becomes more uniform—the higher the temperature, the flatter the output distribution, and the more fine-grained information the student model captures.66% similarUnverifiedAcceptance length and acceptance rate do not have a simple linear relationship: it depends on the product of the probabilities that prefix tokens are accepted one by one.63% similarUnverifiedFor weighted selection, use an unbiased randomInt over the total weight to draw an integer and walk the list, so that each entry's probability is exactly weight / total.62% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1181371API
curl https://kongchang.com/api/v1/knowledge/claims/1181371MCP
get_claim(id=1181371)