Probability Distribution
In the context of decision models, a probability vector obtained by computing conditional log-probabilities over all candidate answers and normalizing the results. This distribution quantifies the model's confidence across individual options and can directly drive programmatic decision logic.
Core Facts
The essence of large language models is learning the conditional probability distribution of what the next word is given the preceding text
When generating each token, LLMs randomly sample from a probability distribution according to a temperature parameter rather than selecting the deterministic output with the highest probability, thereby producing emergent properties.
Timeline (last 90 days)
For multiple-choice questions, the logits over each option token can be normalized via softmax to obtain a probability distribution, where a highly concentrated distribution indicates the model is more confident in the answer
护栏模型通常输出各风险类别的概率分布，开发者可根据业务需要设定不同阈值来平衡精确率与召回率
All Facts (4)
The essence of large language models is learning the conditional probability distribution of what the next word is given the preceding text
80%VerifiedWhen generating each token, LLMs randomly sample from a probability distribution according to a temperature parameter rather than selecting the deterministic output with the highest probability, thereby producing emergent properties.
70%UnverifiedFor multiple-choice questions, the logits over each option token can be normalized via softmax to obtain a probability distribution, where a highly concentrated distribution indicates the model is more confident in the answer
50%Unverified护栏模型通常输出各风险类别的概率分布，开发者可根据业务需要设定不同阈值来平衡精确率与召回率
50%