Problem Formulation

$$$$log(p(y_L / D ; theta) == sum(limits_(i == 1)**m log(p(y_L**i

/ D ; theta))))

Problem Formulation

$$$$log(p(y_s / D ; theta) == sum(limits_(i == 1)**k log(p(w__deriv_i

/ D, w__deriv_1, cdots, w__deriv_(i - 1) ; theta))))

Neural Summarization Model

$$$$f_j**i == tanh(W_(j : j + c - 1) otimes K + b)

Problem Formulation

$$$$log(p(y_L / D ; theta) == sum(limits_(i == 1)**m log(p(y_L**i

/ D ; theta))))

Problem Formulation

$$$$log(p(y_s / D ; theta) == sum(limits_(i == 1)**k log(p(w__deriv_i

/ D, w__deriv_1, cdots, w__deriv_(i - 1) ; theta))))

Neural Summarization Model

$$$$f_j**i == tanh(W_(j : j + c - 1) otimes K + b)

Problem Formulation

$$$$log(p(y_L / D ; theta) == sum(limits_(i == 1)**m log(p(y_L**i

/ D ; theta))))

Problem Formulation

$$$$log(p(y_s / D ; theta) == sum(limits_(i == 1)**k log(p(w__deriv_i

/ D, w__deriv_1, cdots, w__deriv_(i - 1) ; theta))))

Neural Summarization Model

$$$$f_j**i == tanh(W_(j : j + c - 1) otimes K + b)