Unverified50% confidenceFactExact time
The raw scores output by the model are logits, which need to be converted into a probability distribution via the softmax function; temperature is precisely the scaling applied to the logits before softmax: as temperature approaches 0 the distribution becomes sharp (approaching greedy), and as temperature rises the distribution becomes flat.
1
Sources
50%
Confidence
Long-term
Relevance
10/19/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedTemperature is a hyperparameter that controls the randomness of a language model's output; it acts on logits by dividing the logit by temperature T before softmax normalization—when T<1 the distribution sharpens and output becomes more deterministic, when T>1 the distribution flattens and output becomes more random.76% similarUnverifiedThe core of model distillation lies in the application of the temperature softmax function, which softens the probability distribution through the temperature parameter T to reveal similarity relationships between classes71% similarUnverifiedLLM基于Transformer架构的自注意力机制和Softmax概率分布，通过温度参数控制采样随机性，温度趋近0时输出最高概率token，温度升高时引入更多随机性70% similarVerifiedModel distillation softens the teacher model's softmax outputs via temperature scaling, allowing information from low-probability classes to be transferred to the student model.69% similarVerifiedWhen a large language model generates each token, it samples from a probability distribution; the temperature parameter controls the degree of randomness—higher temperature produces more diverse outputs, while lower temperature produces more deterministic ones.68% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1067377API
curl https://kongchang.com/api/v1/knowledge/claims/1067377MCP
get_claim(id=1067377)