Adds the arguments used in management Args: parser: the parser instance Returns: None: This function does not return a value

Adds the arguments used by all the models. Args: parser: the parser instance Returns: None: This function does not return a

Build Worlds-Queries matrix

Flushing print Args: args: arguments kwargs: key-value arguments Returns: None: This function does not return a valu

Compute the coverage of a confusion matrix. Essentially this metric is

Load a CLIP model Parameters ---------- name : str A model name listed by `clip.available_models()`, or the path to a model check

Initialize common components for sampling methods. Args: x_BX: Input tensor w_T_BW: Initial w tensor (can be masked)

Generate and plot a confusion matrix using Matplotlib with normalization. Parameters: y_true (array-like): Ground truth labels.

Initialize method Args: self: instance loss: loss function nr_classes: number of classes Returns

Get tag for the model name Args: args: command line arguments Returns: tag (str): tag for the model name

Create path function, create folder if it does not exists Args: path (str): path value Returns: None: This function does not

Evaluate metrics of the frequentist model Args: model (nn.Module): network loader: dataloader args: command line argument

Creates and returns a continual dataset. :param args: the arguments which contains the hyperparameters :return: the continual dataset

Forward method Args: self: instance x (torch.tensor): input vector Returns: out_dict: output dic

Performs ProbLog inference to retrieve the worlds probability distribution P(w). Works with an arbitrary number of encoded bits (digits).

Prints out the progress bar on the stderr file. Args: i: the current iteration max_iter: the maximum number of iteration

creates save directory if one does not exist save_name: full save path

Build Worlds-Queries matrix

Compute the concept distribution entropy conditioned on the query.

Computes the ECE Args: confs (ndarray): confidence preds (ndarray): predictions labels (ndarray): labels num_bins

Addmnist classification loss Args: out_dict: output dictionary args: command line arguments Returns: loss: loss valu

Addmnist concept match loss Args: out_dict: output dictionary args: command line arguments Returns: loss: loss value

Returns list of utf-8 byte and a corresponding list of unicode strings. The reversible bpe codes work on unicode strings. This means you