Conditional probability
Conditional probability represents uncertainty about one event given information about another. Autoregressive language models factor sequence likelihood into conditional next-token probabilities.
P(A\mid B)=\frac{P(A\cap B)}{P(B)}
Overview. Conditional probability represents uncertainty about one event given information about another.
Context. Autoregressive language models factor sequence likelihood into conditional next-token probabilities.
Claims and evidence
Claim ledger, 2Conditional probability represents uncertainty about one event given information about another.
Autoregressive language models factor sequence likelihood into conditional next-token probabilities.