default

Returns the absolute value of each element in tensor x.

Returns element-wise addition of tensors x and y.

Adds scalar s to each element in tensor x.

Computes the binary cross entropy loss between the prediction and target tensors.

Performs batched matrix multiplication of tensors x and y.

Computes the categorical cross entropy loss between the prediction and target tensors.

Applies ceiling to each element in tensor x.

Clips the elements of tensor x to have a maximum norm of maxNorm.

Clips the elements of tensor x to be within the range [min, max].

Returns element-wise division of tensor x by tensor y.

Divides each element in tensor x by scalar s.

Computes the dot product of tensors x and y.

Derivative of the ELU activation function with an optional alpha parameter.

Exponential Linear Unit (ELU) activation function with an optional alpha parameter.

Computes the exponential of each element in tensor x.

Applies floor to each element in tensor x.

Creates a tensor with the same shape as t filled with the specified value.

Gaussian activation function.

Derivative of the GeLU activation function.

Gaussian Error Linear Unit (GeLU) activation function.

Derivative of the identity activation function.

Identity activation function.

Returns the natural logarithm of each element in tensor x.

Returns the logarithm of tensor x with the specified base.

Logistic activation function with parameters a, b, and c.

Logit activation function.

Derivative of the Leaky ReLU activation function.

Leaky ReLU activation function with an optional alpha parameter.

Computes the Mean Absolute Error (MAE) loss between the prediction and target tensors.

Performs matrix multiplication of tensors x and y.

Returns the element-wise maximum of tensors x and y.

Returns the element-wise maximum between tensor x and scalar s.

Returns the maximum value in tensor x.

Computes the mean of all elements in tensor x.

Computes the mean along the specified axis of tensor x.

Returns the element-wise minimum of tensors x and y.

Returns the element-wise minimum between tensor x and scalar s.

Returns the minimum value in tensor x.

Computes the Mean Squared Error (MSE) loss between the prediction and target tensors.

Returns element-wise multiplication of tensors x and y.

Multiplies each element in tensor x by scalar s.

Returns the negation of tensor x.

Computes the Frobenius norm of tensor x.

Computes the Frobenius norm squared of tensor x.

Creates a tensor with the same shape as x filled with ones.

Permutes the axes of tensor x according to the specified order.

Raises each element in tensor x to the power of s.

Derivative of the ReLU activation function.

Rectified Linear Unit (ReLU) activation function.

Returns the reciprocal of each element in tensor x.

Rounds each element in tensor x to the nearest integer.

Creates a tensor consisting of scalar value s.

Returns the shape of tensor x as a list of dimension sizes.

Derivative of the sigmoid activation function.

Sigmoid activation function.

Returns the sign of each element in tensor x.

Derivative of the softmax activation function.

Softmax activation function.

Returns the square root of each element in tensor x.

Computes the Sum of Squared Errors (SSE) loss between the prediction and target tensors.

Standardizes tensor x along the specified axis.

Computes the standard deviation of tensor x.

Computes the standard deviation along the specified axis of tensor x.

Returns element-wise subtraction of tensor y from tensor x.

Subtracts scalar s from each element in tensor x.

Computes the sum of all elements in tensor x.

Computes the sum along the specified axis of tensor x.

Derivative of the tanh activation function.

Hyperbolic tangent (tanh) activation function.

Transposes tensor x by swapping the specified axes i and j.

Computes the variance of all elements in tensor x.

Computes the variance along the specified axis of tensor x.

Creates a tensor with the same shape as x filled with zeros.