AIOne Jev request reportedly supplied all conditional probability tables for an eight-variable network
A user testing Jev with GTSAM proposes using the model to supply probabilities from knowledge expressed in language. They say two experiments cost 0.03 cents in total.
TLDR
An experimenter describes Jev as returning probabilities for a question’s possible outcomes, and explores using it to supply inputs for graphical-model reasoning engines such as GTSAM. They report that one Jev request supplied every conditional probability table for an eight-variable network. In the replies, another user says they had planned to try the same idea with a different backend; the experimenter says it could probably be adapted quickly.
One Jev request reportedly supplied all conditional probability tables for an eight-variable network
A user testing Jev with GTSAM proposes using the model to supply probabilities from knowledge expressed in language. They say two experiments cost 0.03 cents in total.