| GLM families | glm_families |
| Runs multiple Fisher scoring steps | glm_fit |
| Runs one Fisher scoring step | glm_fit_one_step |
| Runs one Fisher Scoring step | glm_fit_one_step.tensorflow.tensor |
| Runs multiple Fisher scoring steps | glm_fit.tensorflow.tensor |
| Blockwise Initializer | initializer_blockwise |
| Installs TensorFlow Probability | install_tfprobability |
| Masked Autoencoder for Distribution Estimation | layer_autoregressive |
| An autoregressive normalizing flow layer, given a 'layer_autoregressive'. | layer_autoregressive_transform |
| A OneHotCategorical mixture Keras layer from 'k * (1 + d)' params. | layer_categorical_mixture_of_one_hot_categorical |
| 1D convolution layer (e.g. temporal convolution) with Flipout | layer_conv_1d_flipout |
| 1D convolution layer (e.g. temporal convolution). | layer_conv_1d_reparameterization |
| 2D convolution layer (e.g. spatial convolution over images) with Flipout | layer_conv_2d_flipout |
| 2D convolution layer (e.g. spatial convolution over images) | layer_conv_2d_reparameterization |
| 3D convolution layer (e.g. spatial convolution over volumes) with Flipout | layer_conv_3d_flipout |
| 3D convolution layer (e.g. spatial convolution over volumes) | layer_conv_3d_reparameterization |
| Densely-connected layer class with Flipout estimator. | layer_dense_flipout |
| Densely-connected layer class with local reparameterization estimator. | layer_dense_local_reparameterization |
| Densely-connected layer class with reparameterization estimator. | layer_dense_reparameterization |
| Dense Variational Layer | layer_dense_variational |
| Keras layer enabling plumbing TFP distributions through Keras models | layer_distribution_lambda |
| An Independent-Bernoulli Keras layer from prod(event_shape) params | layer_independent_bernoulli |
| An independent Logistic Keras layer. | layer_independent_logistic |
| An independent Normal Keras layer. | layer_independent_normal |
| An independent Poisson Keras layer. | layer_independent_poisson |
| Pass-through layer that adds a KL divergence penalty to the model loss | layer_kl_divergence_add_loss |
| Regularizer that adds a KL divergence penalty to the model loss | layer_kl_divergence_regularizer |
| A mixture distribution Keras layer, with independent logistic components. | layer_mixture_logistic |
| A mixture distribution Keras layer, with independent normal components. | layer_mixture_normal |
| A mixture (same-family) Keras layer. | layer_mixture_same_family |
| A d-variate Multivariate Normal TriL Keras layer from 'd+d*(d+1)/ 2' params | layer_multivariate_normal_tri_l |
| A 'd'-variate OneHotCategorical Keras layer from 'd' params. | layer_one_hot_categorical |
| Variable Layer | layer_variable |
| A Variational Gaussian Process Layer. | layer_variational_gaussian_process |
| Adapts the inner kernel's 'step_size' based on 'log_accept_prob'. | mcmc_dual_averaging_step_size_adaptation |
| Estimate a lower bound on effective sample size for each independent chain. | mcmc_effective_sample_size |
| Runs one step of Hamiltonian Monte Carlo. | mcmc_hamiltonian_monte_carlo |
| Runs one step of Metropolis-adjusted Langevin algorithm. | mcmc_metropolis_adjusted_langevin_algorithm |
| Runs one step of the Metropolis-Hastings algorithm. | mcmc_metropolis_hastings |
| Runs one step of the No U-Turn Sampler | mcmc_no_u_turn_sampler |
| Gelman and Rubin (1992)'s potential scale reduction for chain convergence. | mcmc_potential_scale_reduction |
| Runs one step of the RWM algorithm with symmetric proposal. | mcmc_random_walk_metropolis |
| Runs one step of the Replica Exchange Monte Carlo | mcmc_replica_exchange_mc |
| Runs annealed importance sampling (AIS) to estimate normalizing constants. | mcmc_sample_annealed_importance_chain |
| Implements Markov chain Monte Carlo via repeated 'TransitionKernel' steps. | mcmc_sample_chain |
| Returns a sample from the 'dim' dimensional Halton sequence. | mcmc_sample_halton_sequence |
| Adapts the inner kernel's 'step_size' based on 'log_accept_prob'. | mcmc_simple_step_size_adaptation |
| Runs one step of the slice sampler using a hit and run approach | mcmc_slice_sampler |
| Applies a bijector to the MCMC's state space | mcmc_transformed_transition_kernel |
| Runs one step of Uncalibrated Hamiltonian Monte Carlo | mcmc_uncalibrated_hamiltonian_monte_carlo |
| Runs one step of Uncalibrated Langevin discretized diffusion. | mcmc_uncalibrated_langevin |
| Generate proposal for the Random Walk Metropolis algorithm. | mcmc_uncalibrated_random_walk |
| number of 'params' needed to create a CategoricalMixtureOfOneHotCategorical distribution | params_size_categorical_mixture_of_one_hot_categorical |
| number of 'params' needed to create an IndependentBernoulli distribution | params_size_independent_bernoulli |
| number of 'params' needed to create an IndependentLogistic distribution | params_size_independent_logistic |
| number of 'params' needed to create an IndependentNormal distribution | params_size_independent_normal |
| number of 'params' needed to create an IndependentPoisson distribution | params_size_independent_poisson |
| number of 'params' needed to create a MixtureLogistic distribution | params_size_mixture_logistic |
| number of 'params' needed to create a MixtureNormal distribution | params_size_mixture_normal |
| number of 'params' needed to create a MixtureSameFamily distribution | params_size_mixture_same_family |
| number of 'params' needed to create a MultivariateNormalTriL distribution | params_size_multivariate_normal_tri_l |
| number of 'params' needed to create a OneHotCategorical distribution | params_size_one_hot_categorical |
| A state space model representing a sum of component state space models. | sts_additive_state_space_model |
| Formal representation of an autoregressive model. | sts_autoregressive |
| State space model for an autoregressive process. | sts_autoregressive_state_space_model |
| Build a variational posterior that factors over model parameters. | sts_build_factored_surrogate_posterior |
| Build a loss function for variational inference in STS models. | sts_build_factored_variational_loss |
| Seasonal state space model with effects constrained to sum to zero. | sts_constrained_seasonal_state_space_model |
| Decompose an observed time series into contributions from each component. | sts_decompose_by_component |
| Decompose a forecast distribution into contributions from each component. | sts_decompose_forecast_by_component |
| Formal representation of a dynamic linear regression model. | sts_dynamic_linear_regression |
| State space model for a dynamic linear regression from provided covariates. | sts_dynamic_linear_regression_state_space_model |
| Draw posterior samples using Hamiltonian Monte Carlo (HMC) | sts_fit_with_hmc |
| Construct predictive distribution over future observations | sts_forecast |
| Formal representation of a linear regression from provided covariates. | sts_linear_regression |
| Formal representation of a local level model | sts_local_level |
| State space model for a local level | sts_local_level_state_space_model |
| Formal representation of a local linear trend model | sts_local_linear_trend |
| State space model for a local linear trend | sts_local_linear_trend_state_space_model |
| Compute one-step-ahead predictive distributions for all timesteps | sts_one_step_predictive |
| Initialize from a uniform [-2, 2] distribution in unconstrained space. | sts_sample_uniform_initial_state |
| Formal representation of a seasonal effect model. | sts_seasonal |
| State space model for a seasonal effect. | sts_seasonal_state_space_model |
| Formal representation of a semi-local linear trend model. | sts_semi_local_linear_trend |
| State space model for a semi-local linear trend. | sts_semi_local_linear_trend_state_space_model |
| Formal representation of a smooth seasonal effect model | sts_smooth_seasonal |
| State space model for a smooth seasonal effect | sts_smooth_seasonal_state_space_model |
| Formal representation of a sparse linear regression. | sts_sparse_linear_regression |
| Sum of structural time series components. | sts_sum |
| Computes'Y = g(X) = Abs(X)', element-wise | tfb_absolute_value |
| Affine bijector | tfb_affine |
| ComputesY = g(X; shift, scale) = scale @ X + shift | tfb_affine_linear_operator |
| Maps unconstrained R^n to R^n in ascending order. | tfb_ascending |
| Computes'Y = g(X)' s.t. 'X = g^-1(Y) = (Y - mean(Y)) / std(Y)' | tfb_batch_normalization |
| Bijector which applies a list of bijectors to blocks of a Tensor | tfb_blockwise |
| Bijector which applies a sequence of bijectors | tfb_chain |
| Computes'g(X) = X @ X.T' where 'X' is lower-triangular, positive-diagonal matrix | tfb_cholesky_outer_product |
| Maps the Cholesky factor of M to the Cholesky factor of 'M^{-1}' | tfb_cholesky_to_inv_cholesky |
| Maps unconstrained reals to Cholesky-space correlation matrices. | tfb_correlation_cholesky |
| Computes the cumulative sum of a tensor along a specified axis. | tfb_cumsum |
| Computes'Y = g(X) = DCT(X)', where DCT type is indicated by the type arg | tfb_discrete_cosine_transform |
| Computes'Y=g(X)=exp(X)' | tfb_exp |
| Computes'Y = g(X) = exp(X) - 1' | tfb_expm1 |
| Implements a continuous normalizing flow X->Y defined via an ODE. | tfb_ffjord |
| Transforms unconstrained vectors to TriL matrices with positive diagonal | tfb_fill_scale_tri_l |
| Transforms vectors to triangular | tfb_fill_triangular |
| Returns the forward Bijector evaluation, i.e., 'X = g(Y)'. | tfb_forward |
| Returns the result of the forward evaluation of the log determinant of the Jacobian | tfb_forward_log_det_jacobian |
| Implements the Glow Bijector from Kingma & Dhariwal (2018). | tfb_glow |
| Compute Y = g(X) = 1 - exp(-c * (exp(rate * X) - 1), the Gompertz CDF. | tfb_gompertz_cdf |
| Computes'Y = g(X) = exp(-exp(-(X - loc) / scale))' | tfb_gumbel |
| Compute 'Y = g(X) = exp(-exp(-(X - loc) / scale))', the Gumbel CDF. | tfb_gumbel_cdf |
| Computes'Y = g(X) = X' | tfb_identity |
| Bijector constructed from custom functions | tfb_inline |
| Returns the inverse Bijector evaluation, i.e., 'X = g^{-1}(Y)'. | tfb_inverse |
| Returns the result of the inverse evaluation of the log determinant of the Jacobian | tfb_inverse_log_det_jacobian |
| Bijector which inverts another Bijector | tfb_invert |
| Bijector which applies a Stick Breaking procedure. | tfb_iterated_sigmoid_centered |
| Computes'Y = g(X) = (1 - (1 - X)**(1 / b))**(1 / a)', with X in [0, 1] | tfb_kumaraswamy |
| Computes'Y = g(X) = (1 - (1 - X)**(1 / b))**(1 / a)', with X in [0, 1] | tfb_kumaraswamy_cdf |
| LambertWTail transformation for heavy-tail Lambert W x F random variables. | tfb_lambert_w_tail |
| Masked Autoregressive Density Estimator | tfb_masked_autoregressive_default_template |
| Affine MaskedAutoregressiveFlow bijector | tfb_masked_autoregressive_flow |
| Autoregressively masked dense layer | tfb_masked_dense |
| Computes 'g(L) = inv(L)', where L is a lower-triangular matrix | tfb_matrix_inverse_tri_l |
| Matrix-vector multiply using LU decomposition | tfb_matvec_lu |
| Computes'Y = g(X) = NormalCDF(x)' | tfb_normal_cdf |
| Bijector which maps a tensor x_k that has increasing elements in the last dimension to an unconstrained tensor y_k | tfb_ordered |
| Pads a value to the 'event_shape' of a 'Tensor'. | tfb_pad |
| Permutes the rightmost dimension of a Tensor | tfb_permute |
| Computes'Y = g(X) = (1 + X * c)**(1 / c)', where 'X >= -1 / c' | tfb_power_transform |
| A piecewise rational quadratic spline, as developed in Conor et al.(2019). | tfb_rational_quadratic_spline |
| Compute Y = g(X) = 1 - exp( -(X/scale)**2 / 2 ), X >= 0. | tfb_rayleigh_cdf |
| RealNVP affine coupling layer for vector-valued events | tfb_real_nvp |
| Build a scale-and-shift function using a multi-layer neural network | tfb_real_nvp_default_template |
| A Bijector that computes 'b(x) = 1. / x' | tfb_reciprocal |
| Reshapes the event_shape of a Tensor | tfb_reshape |
| Compute Y = g(X; scale) = scale * X. | tfb_scale |
| Compute Y = g(X; scale) = scale @ X | tfb_scale_matvec_diag |
| Compute Y = g(X; scale) = scale @ X. | tfb_scale_matvec_linear_operator |
| Matrix-vector multiply using LU decomposition. | tfb_scale_matvec_lu |
| Compute Y = g(X; scale) = scale @ X. | tfb_scale_matvec_tri_l |
| Transforms unconstrained vectors to TriL matrices with positive diagonal | tfb_scale_tri_l |
| Compute Y = g(X; shift) = X + shift. | tfb_shift |
| Compute 'Y = g(X) = (1 - exp(-rate * X)) * exp(-c * exp(-rate * X))' | tfb_shifted_gompertz_cdf |
| Computes'Y = g(X) = 1 / (1 + exp(-X))' | tfb_sigmoid |
| Bijector that computes 'Y = sinh(X)'. | tfb_sinh |
| Computes'Y = g(X) = Sinh( (Arcsinh(X) + skewness) * tailweight )' | tfb_sinh_arcsinh |
| Computes Y = g(X) = exp([X 0]) / sum(exp([X 0])) | tfb_softmax_centered |
| Computes 'Y = g(X) = Log[1 + exp(X)]' | tfb_softplus |
| Computes Y = g(X) = X / (1 + |X|) | tfb_softsign |
| Split a 'Tensor' event along an axis into a list of 'Tensor's. | tfb_split |
| Computes'g(X) = X^2'; X is a positive real number. | tfb_square |
| Computes 'Y = tanh(X)' | tfb_tanh |
| Applies a Bijector to the diagonal of a matrix | tfb_transform_diagonal |
| Computes'Y = g(X) = transpose_rightmost_dims(X, rightmost_perm)' | tfb_transpose |
| Computes'Y = g(X) = 1 - exp((-X / scale) ** concentration)' where X >= 0 | tfb_weibull |
| Compute Y = g(X) = 1 - exp((-X / scale) ** concentration), X >= 0. | tfb_weibull_cdf |
| Autoregressive distribution | tfd_autoregressive |
| Batch-Reshaping distribution | tfd_batch_reshape |
| Bates distribution. | tfd_bates |
| Bernoulli distribution | tfd_bernoulli |
| Beta distribution | tfd_beta |
| Beta-Binomial compound distribution | tfd_beta_binomial |
| Binomial distribution | tfd_binomial |
| Blockwise distribution | tfd_blockwise |
| Categorical distribution over integers | tfd_categorical |
| Cauchy distribution with location 'loc' and scale 'scale' | tfd_cauchy |
| Cumulative distribution function. Given random variable X, the cumulative distribution function cdf is: 'cdf(x) := P[X <= x]' | tfd_cdf |
| Chi distribution | tfd_chi |
| Chi Square distribution | tfd_chi2 |
| The CholeskyLKJ distribution on cholesky factors of correlation matrices | tfd_cholesky_lkj |
| Continuous Bernoulli distribution. | tfd_continuous_bernoulli |
| Covariance. | tfd_covariance |
| Computes the (Shannon) cross entropy. | tfd_cross_entropy |
| Scalar 'Deterministic' distribution on the real line | tfd_deterministic |
| Dirichlet distribution | tfd_dirichlet |
| Dirichlet-Multinomial compound distribution | tfd_dirichlet_multinomial |
| Double-sided Maxwell distribution. | tfd_doublesided_maxwell |
| Empirical distribution | tfd_empirical |
| Shannon entropy in nats. | tfd_entropy |
| ExpGamma distribution. | tfd_exp_gamma |
| ExpInverseGamma distribution. | tfd_exp_inverse_gamma |
| ExpRelaxedOneHotCategorical distribution with temperature and logits. | tfd_exp_relaxed_one_hot_categorical |
| Exponential distribution | tfd_exponential |
| The finite discrete distribution. | tfd_finite_discrete |
| Gamma distribution | tfd_gamma |
| Gamma-Gamma distribution | tfd_gamma_gamma |
| Marginal distribution of a Gaussian process at finitely many points. | tfd_gaussian_process |
| Posterior predictive distribution in a conjugate GP regression model. | tfd_gaussian_process_regression_model |
| The Generalized Normal distribution. | tfd_generalized_normal |
| The Generalized Pareto distribution. | tfd_generalized_pareto |
| Geometric distribution | tfd_geometric |
| Scalar Gumbel distribution with location 'loc' and 'scale' parameters | tfd_gumbel |
| Half-Cauchy distribution | tfd_half_cauchy |
| Half-Normal distribution with scale 'scale' | tfd_half_normal |
| Hidden Markov model distribution | tfd_hidden_markov_model |
| Horseshoe distribution | tfd_horseshoe |
| Independent distribution from batch of distributions | tfd_independent |
| InverseGamma distribution | tfd_inverse_gamma |
| Inverse Gaussian distribution | tfd_inverse_gaussian |
| Johnson's SU-distribution. | tfd_johnson_s_u |
| Joint distribution parameterized by named distribution-making functions. | tfd_joint_distribution_named |
| Joint distribution parameterized by named distribution-making functions. | tfd_joint_distribution_named_auto_batched |
| Joint distribution parameterized by distribution-making functions | tfd_joint_distribution_sequential |
| Joint distribution parameterized by distribution-making functions. | tfd_joint_distribution_sequential_auto_batched |
| Computes the Kullback-Leibler divergence. | tfd_kl_divergence |
| Kumaraswamy distribution | tfd_kumaraswamy |
| Laplace distribution with location 'loc' and 'scale' parameters | tfd_laplace |
| Observation distribution from a linear Gaussian state space model | tfd_linear_gaussian_state_space_model |
| LKJ distribution on correlation matrices | tfd_lkj |
| Log cumulative distribution function. | tfd_log_cdf |
| The log-logistic distribution. | tfd_log_logistic |
| Log-normal distribution | tfd_log_normal |
| Log probability density/mass function. | tfd_log_prob |
| Log survival function. | tfd_log_survival_function |
| Logistic distribution with location 'loc' and 'scale' parameters | tfd_logistic |
| The Logit-Normal distribution | tfd_logit_normal |
| Mean. | tfd_mean |
| Mixture distribution | tfd_mixture |
| Mixture (same-family) distribution | tfd_mixture_same_family |
| Mode. | tfd_mode |
| Multinomial distribution | tfd_multinomial |
| Multivariate normal distribution on 'R^k' | tfd_multivariate_normal_diag |
| Multivariate normal distribution on 'R^k' | tfd_multivariate_normal_diag_plus_low_rank |
| Multivariate normal distribution on 'R^k' | tfd_multivariate_normal_full_covariance |
| The multivariate normal distribution on 'R^k' | tfd_multivariate_normal_linear_operator |
| The multivariate normal distribution on 'R^k' | tfd_multivariate_normal_tri_l |
| Multivariate Student's t-distribution on 'R^k' | tfd_multivariate_student_t_linear_operator |
| NegativeBinomial distribution | tfd_negative_binomial |
| Normal distribution with loc and scale parameters | tfd_normal |
| OneHotCategorical distribution | tfd_one_hot_categorical |
| Pareto distribution | tfd_pareto |
| Modified PERT distribution for modeling expert predictions. | tfd_pert |
| The Pixel CNN++ distribution | tfd_pixel_cnn |
| Plackett-Luce distribution over permutations. | tfd_plackett_luce |
| Poisson distribution | tfd_poisson |
| 'PoissonLogNormalQuadratureCompound' distribution | tfd_poisson_log_normal_quadrature_compound |
| The Power Spherical distribution over unit vectors on 'S^{n-1}'. | tfd_power_spherical |
| Probability density/mass function. | tfd_prob |
| ProbitBernoulli distribution. | tfd_probit_bernoulli |
| Quantile function. Aka "inverse cdf" or "percent point function". | tfd_quantile |
| Distribution representing the quantization 'Y = ceiling(X)' | tfd_quantized |
| RelaxedBernoulli distribution with temperature and logits parameters | tfd_relaxed_bernoulli |
| RelaxedOneHotCategorical distribution with temperature and logits | tfd_relaxed_one_hot_categorical |
| Generate samples of the specified shape. | tfd_sample |
| Sample distribution via independent draws. | tfd_sample_distribution |
| The SinhArcsinh transformation of a distribution on (-inf, inf) | tfd_sinh_arcsinh |
| Skellam distribution. | tfd_skellam |
| The uniform distribution over unit vectors on 'S^{n-1}'. | tfd_spherical_uniform |
| Standard deviation. | tfd_stddev |
| Student's t-distribution | tfd_student_t |
| Marginal distribution of a Student's T process at finitely many points | tfd_student_t_process |
| Survival function. | tfd_survival_function |
| A Transformed Distribution | tfd_transformed_distribution |
| Triangular distribution with 'low', 'high' and 'peak' parameters | tfd_triangular |
| The Truncated Cauchy distribution. | tfd_truncated_cauchy |
| Truncated Normal distribution | tfd_truncated_normal |
| Uniform distribution with 'low' and 'high' parameters | tfd_uniform |
| Variance. | tfd_variance |
| Posterior predictive of a variational Gaussian process | tfd_variational_gaussian_process |
| Vector Deterministic Distribution | tfd_vector_deterministic |
| VectorDiffeomixture distribution | tfd_vector_diffeomixture |
| The vectorization of the Exponential distribution on 'R^k' | tfd_vector_exponential_diag |
| The vectorization of the Exponential distribution on 'R^k' | tfd_vector_exponential_linear_operator |
| The vectorization of the Laplace distribution on 'R^k' | tfd_vector_laplace_diag |
| The vectorization of the Laplace distribution on 'R^k' | tfd_vector_laplace_linear_operator |
| The (diagonal) SinhArcsinh transformation of a distribution on 'R^k' | tfd_vector_sinh_arcsinh_diag |
| The von Mises distribution over angles | tfd_von_mises |
| The von Mises-Fisher distribution over unit vectors on 'S^{n-1}' | tfd_von_mises_fisher |
| The Weibull distribution with 'concentration' and 'scale' parameters. | tfd_weibull |
| The matrix Wishart distribution on positive definite matrices | tfd_wishart |
| The matrix Wishart distribution on positive definite matrices | tfd_wishart_linear_operator |
| The matrix Wishart distribution parameterized with Cholesky factors. | tfd_wishart_tri_l |
| Zipf distribution | tfd_zipf |
| Handle to the 'tensorflow_probability' module | tfp |
| TensorFlow Probability Version | tfp_version |
| The Amari-alpha Csiszar-function in log-space | vi_amari_alpha |
| The Arithmetic-Geometric Csiszar-function in log-space | vi_arithmetic_geometric |
| The chi-square Csiszar-function in log-space | vi_chi_square |
| Use VIMCO to lower the variance of the gradient of csiszar_function(Avg(logu)) | vi_csiszar_vimco |
| Calculates the dual Csiszar-function in log-space | vi_dual_csiszar_function |
| Fit a surrogate posterior to a target (unnormalized) log density | vi_fit_surrogate_posterior |
| The Jeffreys Csiszar-function in log-space | vi_jeffreys |
| The Jensen-Shannon Csiszar-function in log-space | vi_jensen_shannon |
| The forward Kullback-Leibler Csiszar-function in log-space | vi_kl_forward |
| The reverse Kullback-Leibler Csiszar-function in log-space | vi_kl_reverse |
| The log1p-abs Csiszar-function in log-space | vi_log1p_abs |
| The Modified-GAN Csiszar-function in log-space | vi_modified_gan |
| Monte-Carlo approximation of an f-Divergence variational loss | vi_monte_carlo_variational_loss |
| The Pearson Csiszar-function in log-space | vi_pearson |
| The Squared-Hellinger Csiszar-function in log-space | vi_squared_hellinger |
| Symmetrizes a Csiszar-function in log-space | vi_symmetrized_csiszar_function |
| The T-Power Csiszar-function in log-space | vi_t_power |
| The Total Variation Csiszar-function in log-space | vi_total_variation |
| The Triangular Csiszar-function in log-space | vi_triangular |
