Unverified50% confidenceFactExact time
Acceptance 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.
1
Sources
50%
Confidence
Long-term
Relevance
11/3/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedConfidence is typically represented by the product of per-token probabilities or the sum of their logarithms.68% similarUnverifiedWithin certain parameter regions, averaging does not always reduce fluctuations, which differs from the intuition in convex optimization.66% similarUnverifiedThe over-refusal rate and the harmful-output rate exhibit a trade-off during parameter tuning, which in machine learning is known as the precision-recall trade-off.65% similarUnverifiedCriterion pass rate refers to the proportion of a set of example model responses in which a candidate criterion is judged to be satisfied; a pass rate that is too high or too low indicates that the criterion lacks discriminative power or is misaligned.65% similarUnverifiedThe standard practice aggregates all heterogeneous harmful prompts into a single uniform mean vector, which may conflate distinct refusal-related phenomena.64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1129371API
curl https://kongchang.com/api/v1/knowledge/claims/1129371MCP
get_claim(id=1129371)