Structured Data Models Structured Data Models Structured Data Models
/

Getting Started

  • Installation
  • Overview

Guides

  • Table Semantics
  • In-Context Learning
  • Data Processing

API

  • sdm
    • Stype
    • infer_stypes
    • Task
  • sdm.tensor
    • VarLenTensor
    • StringTensor
    • NullableTensor
    • CategoricalTensor
    • ColumnarTensor
    • TableTensor
  • sdm.relational
    • Relationship
    • RelationalData
    • TaskLink
    • RelatedTables
    • RelationalSampler
  • sdm.processing
    • Processor
    • InvertibleMixin
    • EnsembleProcessor
    • EnsembleInvertibleMixin
    • Recipe
    • Identity
    • Callable
    • Sequential
    • DropStypes
    • StypeDispatch
    • TaskDispatch
    • TableDispatch
    • EnsembleProcessorAdapter
    • Choice
    • ToNumerical
    • ShuffleColumns
    • SelectColumns
    • Cast
    • Clip
    • ClipQuantiles
    • ClipSigma
    • ClipSoft
    • ImputeMean
    • PowerTransform
    • QuantileTransform
    • RankGaussian
    • Standardize
    • RobustScale
    • FlipSign
    • DropConstantColumns
    • PCA
    • RandomProjection
    • AlignCategories
    • ShuffleCategories
    • ImputeMode
    • AddCategoryCounts
    • TFIDF
    • SentenceTransformer
    • AddCalendarFields
    • AverageEstimators
    • Softmax
    • SortQuantiles
  • sdm.nn
    • RotaryEmbedding
    • SwiGLU
    • SoftplusScale
    • RMSNorm
    • QueryScaling
    • QASSMax
    • LogScale
    • GatedLogScale
    • SDPA
    • Attention
    • TransformerBlock
    • InducedTransformerBlock
  • sdm.models
    • ICLModel
    • TabICLv2
    • TabFM
    • KumoTabular
    • KumoRelational
    • TimesFM3
  • sdm.explain
    • ICLExplainer
    • GradientExplainer
  • sdm.evaluation
    • to_class_indices
    • to_binary_class
  1. Structured Data Models /
  2. sdm.explain

sdm.explain#

ICLExplainer

GradientExplainer

Return gradients showing how small input changes affect predictions.
Previous
TimesFM3
Next
ICLExplainer