KernelSHAP

Bases: MultitaskExplainerMixin, AttributionExplainer

KernelSHAP is a model‑agnostic Shapley value estimator that interprets predictions by computing Shapley values through a weighted linear regression in the space of feature coalitions.

By unifying ideas from LIME and Shapley value theory, KernelSHAP provides additive feature attributions with strong consistency guarantees.

Reference: Lundberg and Lee (2017). A Unified Approach to Interpreting Model Predictions. Paper

Examples:

>>> from interpreto import Granularity, KernelShap
>>> from interpreto.attributions import InferenceModes
>>> method = KernelShap(model, tokenizer, batch_size=4,
>>>                     inference_mode=InferenceModes.SOFTMAX,
>>>                     n_perturbations=20,
>>>                     granularity=Granularity.WORD)
>>> explanations = method(text)

Parameters:

Name	Type	Description	Default
model	PreTrainedModel	model to explain	required
tokenizer	PreTrainedTokenizer	Hugging Face tokenizer associated with the model	required
batch_size	int	batch size for the attribution method	4
granularity	Granularity	granularity level of the perturbations (token, word, sentence, etc.)	WORD
inference_mode	Callable[[Tensor], Tensor]	The mode used for inference. It can be either one of LOGITS, SOFTMAX, or LOG_SOFTMAX. Use InferenceModes to choose the appropriate mode.	LOGITS
n_perturbations	int	the number of perturbations to generate	1000
distance_function	DistancesFromMaskProtocol	distance function used to compute weights of perturbed samples in the linear model training.	required
similarity_kernel	SimilarityKernelProtocol	similarity kernel used to compute weights of perturbed samples in the linear model training.	required
kernel_width	float | Callable	kernel width used in the similarity_kernel	required
device	device	device on which the attribution method will be run	None