What is the difference between a logit and a probability?
A logit is an unnormalized score produced by the model. A normalization operation such as softmax converts the set of scores into probabilities.
Explore This Question
Search for additional explanations using these tools:
More FAQs in How does an LLM predict the next token in a sequence?
Not necessarily. Tokens can be shorter than words or include punctuation and other text units. The tokenizer determines which units the model predicts.
The prediction depends on the broader available context, including earlier tokens and instructions. Different contexts can give the same word different roles and meanings.
Still have questions?
Browse all our FAQs or reach out to our support team