Verified65% confidenceFactExact time
The Softmax formula with temperature parameter T is P(xᵢ) = exp(zᵢ/T) / Σexp(zⱼ/T); when T→0 it degenerates to argmax, and when T→∞ it approaches a uniform distribution
3
Sources
65%
Confidence
Long-term
Relevance
7/20/2026
First Seen
Sources
Related Claims
Unverified温度参数（Temperature）的数学本质是对Softmax函数输出的概率分布进行缩放，趋近于0时模型进行贪心解码71% similarUnverified采样温度T=1时保持原始概率分布，T→0时输出趋于确定性，T>1时分布更平坦、输出更随机69% similarVerifiedThe large model temperature parameter adjusts the flatness of the output probability distribution by dividing logits by the temperature value before applying softmax normalization.68% similarVerifiedThe temperature parameter scales the Logits vector, and when temperature is 0, it degenerates into greedy decoding.67% similarVerifiedThe Temperature parameter acts on the softmax function output: when below 1, the probability distribution becomes sharper; when above 1, the distribution becomes flatter.67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/569077API
curl https://kongchang.com/api/v1/knowledge/claims/569077MCP
get_claim(id=569077)