torch_concepts.nn.GraphConceptBottleneckModel

class GraphConceptBottleneckModel(*args, lightning: bool = False, **kwargs)

Linear concept bottleneck over a DAG: root concepts encoded from the latent, internal concepts predicted from their parent concepts.

Parameters:

• input_size (int) – Dimensionality of input features (after the backbone, if any).

• annotations (Annotations) – Concept annotations (labels, cardinalities, types).

• graph (ConceptGraph) – Directed acyclic graph over the concepts (node names must match labels).

• inference – Inference engine configuration (see ConceptBottleneckModel).

• inference_kwargs – Inference engine configuration (see ConceptBottleneckModel).

• train_inference – Inference engine configuration (see ConceptBottleneckModel).

• train_inference_kwargs – Inference engine configuration (see ConceptBottleneckModel).

• lightning (bool, default False) – If True, adds Lightning training capabilities.

• **kwargs – Forwarded to BaseModel (e.g. backbone, latent_size).

__init__(input_size: int, annotations: ~torch_concepts.annotations.Annotations, graph: ~torch_concepts.concept_graph.ConceptGraph, inference: ~torch_concepts.nn.modules.mid.inference.base.BaseInference | None = <class 'torch_concepts.nn.modules.mid.inference.torch.deterministic.DeterministicInference'>, inference_kwargs: dict | None = None, train_inference: ~torch_concepts.nn.modules.mid.inference.base.BaseInference | None = None, train_inference_kwargs: dict | None = None, lightning: bool = False, **kwargs)

Initialize internal Module state, shared by both nn.Module and ScriptModule.

Methods

__init__(input_size, annotations, graph[, ...])	Initialize internal Module state, shared by both nn.Module and ScriptModule.
add_module(name, module)	Add a child module to the current module.
apply(fn)	Apply fn recursively to every submodule (as returned by .children()) as well as self.
bfloat16()	Casts all floating point parameters and buffers to bfloat16 datatype.
buffers([recurse])	Return an iterator over module buffers.
build_embedding_encoder(n_embeddings)	Build the latent → embeddings layer (n_embeddings of embedding_size).
build_encoder(in_embeddings, out_concepts)	Build the layer encoding a root concept from its latent/embedding.
build_predictor(in_concepts, in_embeddings, ...)	Build the layer predicting an internal concept from its parents.
build_query(ground_truth)	Build the full-observation query that fills every concept's tensor.
children()	Return an iterator over immediate children modules.
compile(*args, **kwargs)	Compile this Module's forward using torch.compile().
cpu()	Move all model parameters and buffers to the CPU.
cuda([device])	Move all model parameters and buffers to the GPU.
dist_kwargs_of(name)	Distribution keyword arguments this model uses for concept name.
distribution_of(name)	Distribution class this model uses for concept name (by its type).
double()	Casts all floating point parameters and buffers to double datatype.
eval()	Set the module in evaluation mode.
extra_repr()	Return the extra representation of the module.
float()	Casts all floating point parameters and buffers to float datatype.
forward(query[, evidence, input])	Unified forward pass for all inference engines.
get_buffer(target)	Return the buffer given by target if it exists, otherwise throw an error.
get_extra_state()	Return any extra state to include in the module's state_dict.
get_parameter(target)	Return the parameter given by target if it exists, otherwise throw an error.
get_submodule(target)	Return the submodule given by target if it exists, otherwise throw an error.
half()	Casts all floating point parameters and buffers to half datatype.
ipu([device])	Move all model parameters and buffers to the IPU.
load_state_dict(state_dict[, strict, assign])	Copy parameters and buffers from state_dict into this module and its descendants.
modules([remove_duplicate])	Return an iterator over all modules in the network.
mtia([device])	Move all model parameters and buffers to the MTIA.
named_buffers([prefix, recurse, ...])	Return an iterator over module buffers, yielding both the name of the buffer as well as the buffer itself.
named_children()	Return an iterator over immediate children modules, yielding both the name of the module as well as the module itself.
named_modules([memo, prefix, remove_duplicate])	Return an iterator over all modules in the network, yielding both the name of the module as well as the module itself.
named_parameters([prefix, recurse, ...])	Return an iterator over module parameters, yielding both the name of the parameter as well as the parameter itself.
parameters([recurse])	Return an iterator over module parameters.
plate_compatible_levels(axis_annotation, graph)	Flag, per graph level, whether its concepts can share a plate.
prepare_target(target)	Prepare ground truth labels for loss/metrics.
register_backward_hook(hook)	Register a backward hook on the module.
register_buffer(name, tensor[, persistent])	Add a buffer to the module.
register_forward_hook(hook, *[, prepend, ...])	Register a forward hook on the module.
register_forward_pre_hook(hook, *[, ...])	Register a forward pre-hook on the module.
register_full_backward_hook(hook[, prepend])	Register a backward hook on the module.
register_full_backward_pre_hook(hook[, prepend])	Register a backward pre-hook on the module.
register_load_state_dict_post_hook(hook)	Register a post-hook to be run after module's load_state_dict() is called.
register_load_state_dict_pre_hook(hook)	Register a pre-hook to be run before module's load_state_dict() is called.
register_module(name, module)	Alias for add_module().
register_parameter(name, param)	Add a parameter to the module.
register_state_dict_post_hook(hook)	Register a post-hook for the state_dict() method.
register_state_dict_pre_hook(hook)	Register a pre-hook for the state_dict() method.
requires_grad_([requires_grad])	Change if autograd should record operations on parameters in this module.
set_extra_state(state)	Set extra state contained in the loaded state_dict.
set_submodule(target, module[, strict])	Set the submodule given by target if it exists, otherwise throw an error.
setup_inference(inference[, ...])	Instantiate and store the eval/train inference engines.
share_memory()	See torch.Tensor.share_memory_().
state_dict(*args[, destination, prefix, ...])	Return a dictionary containing references to the whole state of the module.
to(*args, **kwargs)	Move and/or cast the parameters and buffers.
to_empty(*, device[, recurse])	Move the parameters and buffers to the specified device without copying storage.
train([mode])	Set the module in training mode.
type(dst_type)	Casts all parameters and buffers to dst_type.
xpu([device])	Move all model parameters and buffers to the XPU.
zero_grad([set_to_none])	Reset gradients of all model parameters.

Attributes

T_destination
backbone	The backbone mapping raw input to the latent representation.
call_super_init
dump_patches
embedding_size	Per-concept embedding width (set by subclasses that use embeddings).
inference	Return the active inference engine based on train/eval mode.
internal_embeddings	Internal concepts get their own embedding consumed by the predictor.
param_for_discrete_var	Distribution parameter used for discrete variables — "logits" or "probs".
source_embeddings	Root concepts are decoded from a per-concept embedding (not the raw latent).
supported_concept_types
variable_dist_kwargs	Default keyword arguments per distribution class (e.g. relaxation temperature).
variable_distributions	which distribution this model uses for each concept type ('binary' / 'categorical' / 'continuous').
training