NTNInteraction

class NTNInteraction(activation: str | Module | type[Module] | None = None, activation_kwargs: Mapping[str, Any] | None = None)[source]

Bases: FunctionalInteraction[Tensor, tuple[Tensor, Tensor, Tensor, Tensor, Tensor], Tensor]

A stateful module for the NTN interaction function.

See also

pykeen.nn.functional.ntn_interaction()

Initialize NTN with the given non-linear activation function.

Parameters:

  • activation (HintOrType[nn.Module]) – A non-linear activation function. Defaults to the hyperbolic tangent torch.nn.Tanh if None, otherwise uses the pykeen.utils.activation_resolver for lookup.

  • activation_kwargs (Mapping[str, Any] | None) – If the activation is passed as a class, these keyword arguments are used during its instantiation.

Attributes Summary

relation_shape

The symbolic shapes for relation representations

Methods Summary

func(t, w, vh, vt, b, u, activation)

Evaluate the NTN interaction function.

Attributes Documentation

relation_shape: Sequence[str] = ('kdd', 'kd', 'kd', 'k', 'k')

The symbolic shapes for relation representations

Methods Documentation

func(t: Tensor, w: Tensor, vh: Tensor, vt: Tensor, b: Tensor, u: Tensor, activation: Module) → Tensor

Evaluate the NTN interaction function.

\[f(h,r,t) = u_r^T act(h W_r t + V_r h + V_r' t + b_r)\]
Parameters:

  • h (Tensor) – shape: (*batch_dims, dim) The head representations.

  • w (Tensor) – shape: (*batch_dims, k, dim, dim) The relation specific transformation matrix W_r.

  • vh (Tensor) – shape: (*batch_dims, k, dim) The head transformation matrix V_h.

  • vt (Tensor) – shape: (*batch_dims, k, dim) The tail transformation matrix V_h.

  • b (Tensor) – shape: (*batch_dims, k) The relation specific offset b_r.

  • u (Tensor) – shape: (*batch_dims, k) The relation specific final linear transformation b_r.

  • t (Tensor) – shape: (*batch_dims, dim) The tail representations.

  • activation (Module) – The activation function.

Returns:

shape: batch_dims The scores.

Return type:

Tensor