Supported aten operators

Note

This table and page are auto generated by docs/contributing/generate_support_page.py and reflect the PyTorch reference docs at the time of generation. Targeted torch version: 2.13. Generated on 01 Aug 2026. The ONNX section is not scraped from PyTorch's docs: it is measured by actually exporting one module per operator with torch 2.13.0 at opset 18 (see that section for how to read its columns).

Warning

Take these results with a grain of salt: many of the listed operators never appear in the torch IR graph that torch_to_nnef traces (they get remapped to more generic ops upstream), and some uncommon operators are rare in real models so support may be lacking even when marked unsupported. SONOS maintains operators on a per-need basis, and contributions are always welcome see how.

'is core' column refers to this PyTorch IR documentation page (the page for torch 2.13 was emptied upstream; falling back to torch 2.9 which is the last version that still enumerates the core ATen IR).

We filter out 'backward' and 'sym' operators from the listing since they are unwanted in an inference engine. In-place operations are merged with their non-inplace counterparts since that distinction is an inference implementation detail.

We also exclude a long tail of identifiers that the aten::* source-grep picks up but that we cannot tie to an attributable inference operator measurement (see the full list at the bottom of this page). This trims the unsupported column to the names where a torch_to_nnef emitter (or a deliberate no-op map) would actually be meaningful.

Total matched operators in torch_to_nnef compared to:

  • core PyTorch opset:

130/138

  • and support from full aten:::

401/555

(total operators listed as supported by torch_to_nnef being 455)

export&runaten namealiasescan in-placeis core
✅absabsolute✅✅
✅acosarccos✅✅
✅acosharccosh✅✅
✅adaptive_avg_pool1d❌✅
✅add✅✅
✅addmm✅✅
✅alias❌✅
✅amax❌✅
✅amin❌✅
✅any❌✅
✅arange❌✅
✅argmax❌✅
✅argmin❌✅
❌as_strided✅✅
✅asinarcsin✅✅
✅asinharcsinh✅✅
✅atanarctan✅✅
✅atan2arctan2✅✅
✅atanharctanh✅✅
✅avg_pool1d❌✅
✅avg_pool2d❌✅
✅avg_pool3d❌✅
✅bitwise_and✅✅
✅bitwise_not✅✅
✅bitwise_or✅✅
✅bitwise_xor✅✅
✅bmm❌✅
✅catconcat, concatenate❌✅
✅ceil✅✅
✅clampclip✅✅
✅clone❌✅
✅col2im❌✅
✅constant_pad_nd❌✅
✅convolution❌✅
✅copy✅✅
✅cos✅✅
✅cosh✅✅
✅cumsum✅✅
✅diagonal❌✅
✅divdivide, true_divide✅✅
✅elu✅✅
✅embedding❌✅
❌empty❌✅
❌empty_strided❌✅
✅eq✅✅
✅erfspecial_erf✅✅
✅exp✅✅
✅expand❌✅
✅expm1special_expm1✅✅
✅fill✅✅
✅flip❌✅
✅floor✅✅
✅fmod✅✅
✅full❌✅
✅full_like❌✅
✅gather❌✅
✅gegreater_equal✅✅
✅gelu✅✅
✅grid_sampler_2d❌✅
✅gtgreater✅✅
✅hardtanh✅✅
✅index❌✅
✅index_put✅✅
✅index_select❌✅
✅isinf❌✅
✅isnan❌✅
✅leless_equal✅✅
✅leaky_relu✅✅
✅log✅✅
✅log10✅✅
✅log1pspecial_log1p✅✅
✅log2✅✅
✅logical_and✅✅
✅logical_not✅✅
✅logical_or✅✅
✅logical_xor✅✅
✅ltless✅✅
❌masked_scatter✅✅
✅max❌✅
✅max_pool2d_with_indices❌✅
✅max_pool3d_with_indices❌✅
✅maximum❌✅
✅mean❌✅
✅min❌✅
✅minimum❌✅
✅mm❌✅
✅mulmultiply✅✅
✅native_dropout❌✅
✅native_group_norm❌✅
✅native_layer_norm❌✅
✅nenot_equal✅✅
✅negnegative✅✅
❌nonzero❌✅
✅permute❌✅
✅pow✅✅
✅prod❌✅
❌rand❌✅
❌randn❌✅
❌randperm❌✅
✅reciprocal✅✅
✅reflection_pad1d❌✅
✅reflection_pad2d❌✅
✅reflection_pad3d❌✅
✅relu✅✅
✅remainder✅✅
✅repeat❌✅
✅replication_pad2d❌✅
✅replication_pad3d❌✅
✅roundspecial_round✅✅
✅rsqrt✅✅
✅scalar_tensor❌✅
✅scatter✅✅
✅scatter_add✅✅
✅scatter_reduce✅✅
✅select❌✅
✅select_scatter❌✅
✅sigmoidspecial_expit✅✅
✅sign✅✅
✅sin✅✅
✅sinh✅✅
✅slice❌✅
✅slice_scatter❌✅
✅sort❌✅
✅split_with_sizes❌✅
✅sqrt✅✅
✅squeeze✅✅
✅subsubtract✅✅
✅sum❌✅
✅tan✅✅
✅tanh✅✅
✅topk❌✅
✅truncfix✅✅
✅unsqueeze✅✅
✅upsample_bilinear2d❌✅
✅upsample_nearest2d❌✅
✅var❌✅
✅view❌✅
✅where❌✅
✅adaptive_avg_pool2d❌-
✅adaptive_avg_pool3d❌-
✅adaptive_max_pool1d❌-
✅adaptive_max_pool2d❌-
✅adaptive_max_pool3d❌-
✅addbmm✅-
✅addcdiv❌-
✅addcmul❌-
✅addmv✅-
✅addr✅-
✅affine_grid_generator❌-
✅all❌-
✅alpha_dropout✅-
✅aminmax❌-
✅angle❌-
✅argsort❌-
❌argwhere❌-
✅atleast_1d❌-
✅atleast_2d❌-
✅atleast_3d❌-
✅baddbmm✅-
✅bartlett_window❌-
✅batch_norm❌-
❌bernoulli✅-
✅bilinear❌-
✅binary_cross_entropy❌-
✅binary_cross_entropy_with_logits❌-
❌bincount❌-
❌binomial❌-
✅bitwise_left_shift✅-
✅bitwise_right_shift✅-
✅blackman_window❌-
✅block_diag❌-
✅broadcast_tensors❌-
✅broadcast_to❌-
✅bucketize❌-
✅cartesian_prod❌-
❌cauchy✅-
✅cdist❌-
✅celu❌-
✅chain_matmul❌-
❌chalf❌-
✅channel_shuffle❌-
❌cholesky❌-
❌cholesky_inverse❌-
❌cholesky_solve❌-
✅chunk❌-
✅clamp_max✅-
✅clamp_min✅-
✅column_stack❌-
❌combinations❌-
✅complex❌-
✅conj❌-
✅conj_physical✅-
✅contiguous❌-
✅conv1d❌-
✅conv2d❌-
✅conv3d❌-
✅conv_tbc❌-
✅conv_transpose1d❌-
✅conv_transpose2d❌-
✅conv_transpose3d❌-
❌copy_sparse_to_sparse_❌-
✅copysign✅-
❌corrcoef❌-
✅cosine_embedding_loss❌-
✅cosine_similarity❌-
✅count_nonzero❌-
✅cross❌-
✅cross_entropy_loss❌-
❌ctc_loss❌-
✅cummax❌-
✅cummin❌-
✅cumprod✅-
❌cumulative_trapezoid❌-
✅deg2rad❌-
✅dequantize❌-
✅detach✅-
✅diag❌-
✅diag_embed❌-
✅diagflat❌-
✅diff❌-
✅digammaspecial_digamma, special_psi✅-
✅dist❌-
✅dot❌-
✅dropout✅-
✅dsplit❌-
✅dstack❌-
✅einsum❌-
✅embedding_bag❌-
❌embedding_renorm_❌-
✅empty_like❌-
❌empty_permuted❌-
❌empty_quantized❌-
✅erfcspecial_erfc✅-
❌erfinvspecial_erfinv✅-
✅exp2special_exp2✅-
✅expand_as❌-
❌exponential✅-
✅eye❌-
✅feature_alpha_dropout✅-
✅feature_dropout✅-
✅fft_fftfreq❌-
❌fft_ihfft2❌-
❌fft_ihfftn❌-
❌fft_irfftn❌-
✅fft_rfftfreq❌-
❌fft_rfftn❌-
✅flatten❌-
❌flatten_dense_tensors❌-
✅fliplr❌-
✅flipud❌-
✅float_power✅-
✅floor_divide✅-
✅fmax❌-
✅fmin❌-
✅frac❌-
❌fractional_max_pool2d❌-
❌fractional_max_pool3d❌-
✅frexp❌-
✅frobenius_norm❌-
❌gamma❌-
❌gcd✅-
❌geometric✅-
❌geqrf❌-
✅glu❌-
✅grid_sampler❌-
✅grid_sampler_3d❌-
✅group_norm❌-
✅gru❌-
✅hamming_window❌-
✅hann_window❌-
✅hardshrink✅-
✅hardsigmoid✅-
✅hardswish✅-
✅heaviside❌-
✅hinge_embedding_loss❌-
❌histc❌-
❌histogram❌-
❌histogramdd❌-
✅hsplit❌-
✅hstack❌-
✅huber_loss❌-
✅hypot✅-
✅i0special_i0✅-
❌igammaspecial_gammainc✅-
❌igammacspecial_gammaincc✅-
✅im2col❌-
✅imag❌-
✅index_add❌-
✅index_copy❌-
✅index_fill❌-
❌index_put_impl_❌-
❌index_reduce✅-
✅inner❌-
✅instance_norm❌-
✅isclose❌-
✅isfinite❌-
❌isin❌-
✅isneginf❌-
✅isposinf❌-
❌isreal❌-
✅istft❌-
✅kaiser_window❌-
✅kl_div❌-
✅kron❌-
❌kthvalue❌-
✅l1_loss❌-
✅layer_norm❌-
❌lcm✅-
✅ldexp✅-
✅lerp✅-
✅lgamma✅-
❌linalg_cond❌-
✅linalg_cross❌-
❌linalg_detdet❌-
✅linalg_diagonal❌-
❌linalg_eig❌-
❌linalg_eigh❌-
❌linalg_eigvals❌-
❌linalg_eigvalsh❌-
❌linalg_householder_productorgqr❌-
❌linalg_invinverse❌-
❌linalg_ldl_solve❌-
❌linalg_lstsq❌-
❌linalg_lu❌-
❌linalg_lu_solve❌-
❌linalg_matrix_expmatrix_exp❌-
✅linalg_matrix_norm❌-
❌linalg_matrix_powermatrix_power❌-
❌linalg_matrix_rank❌-
✅linalg_norm❌-
❌linalg_pinv❌-
❌linalg_qr❌-
❌linalg_slogdet❌-
❌linalg_solve❌-
❌linalg_solve_triangular❌-
❌linalg_svd❌-
❌linalg_tensorinv❌-
❌linalg_tensorsolve❌-
✅linalg_vector_norm❌-
✅linear❌-
✅linspace❌-
❌log_normal✅-
✅log_sigmoid❌-
✅log_softmaxspecial_log_softmax❌-
✅logaddexp❌-
✅logaddexp2❌-
✅logcumsumexp❌-
❌logdet❌-
✅logitspecial_logit✅-
✅logspace❌-
✅logsumexpspecial_logsumexp❌-
✅lstm❌-
❌lu_solve❌-
❌lu_unpack❌-
✅mHadjoint❌-
✅mT❌-
✅margin_ranking_loss❌-
✅masked_fill✅-
❌masked_select❌-
✅matmullinalg_matmul❌-
✅matrix_H❌-
✅max_pool1d❌-
✅max_pool1d_with_indices❌-
✅max_pool2d❌-
✅max_pool3d❌-
❌max_unpool2d❌-
❌max_unpool3d❌-
❌median❌-
✅meshgrid❌-
✅mish❌-
❌mode❌-
✅movedimmoveaxis❌-
✅mse_loss❌-
❌msort❌-
❌multi_margin_loss❌-
❌multilabel_margin_loss❌-
❌multinomial❌-
✅mv❌-
✅mvlgammaspecial_multigammaln❌-
✅nan_to_num✅-
✅nanmean❌-
❌nanmedian❌-
❌nanquantile❌-
✅nansum❌-
✅narrow❌-
✅native_channel_shuffle❌-
✅new_empty❌-
❌new_empty_strided❌-
✅new_full❌-
✅new_ones❌-
✅new_zeros❌-
❌nextafter✅-
✅nll_loss❌-
✅nll_loss2d❌-
✅nll_loss_nd❌-
❌nonzero_numpy❌-
❌nonzero_static❌-
✅norm❌-
❌normal✅-
❌nuclear_norm❌-
✅numpy_T❌-
✅one_hot❌-
✅ones❌-
✅ones_like❌-
❌ormqr❌-
✅outerger❌-
✅pad❌-
❌pad_sequence❌-
✅pairwise_distance❌-
✅pdist❌-
✅pixel_shuffle❌-
✅pixel_unshuffle❌-
❌poisson❌-
❌poisson_nll_loss❌-
✅polar❌-
❌polygammaspecial_polygamma✅-
✅positive❌-
✅prelu❌-
❌put✅-
❌qr❌-
❌quantile❌-
❌quantize_per_channel❌-
✅quantize_per_tensor❌-
❌quantize_per_tensor_dynamic❌-
❌quantized_batch_norm❌-
❌quantized_max_pool1d❌-
❌quantized_max_pool2d❌-
❌quantized_max_pool3d❌-
✅rad2deg❌-
❌rand_like❌-
❌randint❌-
❌randint_like❌-
❌randn_like❌-
❌random✅-
❌range❌-
✅ravel❌-
✅real❌-
✅relu6✅-
✅renorm❌-
✅repeat_interleave❌-
✅replication_pad1d❌-
✅reshape❌-
✅reshape_as❌-
❌resize_as_sparse_❌-
✅resolve_conj❌-
✅resolve_neg❌-
✅rnn_relu❌-
✅rnn_tanh❌-
✅roll❌-
✅rot90❌-
❌rrelu❌-
❌rrelu_with_noise✅-
✅rsub❌-
✅scaled_dot_product_attention❌-
✅searchsorted❌-
❌segment_reduce❌-
✅selu✅-
❌set_❌-
✅sgn❌-
✅signbit❌-
✅silu✅-
✅sincspecial_sinc❌-
✅smooth_l1_loss❌-
✅soft_margin_loss❌-
✅softmaxspecial_softmax❌-
✅softplus❌-
✅softshrink❌-
❌special_airy_ai❌-
❌special_bessel_j0❌-
❌special_bessel_j1❌-
❌special_bessel_y0❌-
❌special_bessel_y1❌-
❌special_chebyshev_polynomial_t❌-
❌special_chebyshev_polynomial_u❌-
❌special_chebyshev_polynomial_v❌-
❌special_chebyshev_polynomial_w❌-
✅special_entr❌-
❌special_erfcx❌-
❌special_hermite_polynomial_h❌-
❌special_hermite_polynomial_he❌-
✅special_i0e❌-
✅special_i1❌-
✅special_i1e❌-
❌special_laguerre_polynomial_l❌-
❌special_legendre_polynomial_p❌-
❌special_log_ndtr❌-
❌special_modified_bessel_i0❌-
❌special_modified_bessel_i1❌-
❌special_modified_bessel_k0❌-
❌special_modified_bessel_k1❌-
❌special_ndtr❌-
❌special_ndtri❌-
❌special_scaled_modified_bessel_k0❌-
❌special_scaled_modified_bessel_k1❌-
❌special_shifted_chebyshev_polynomial_t❌-
❌special_shifted_chebyshev_polynomial_u❌-
❌special_shifted_chebyshev_polynomial_v❌-
❌special_shifted_chebyshev_polynomial_w❌-
❌special_spherical_bessel_j0❌-
✅special_xlog1py❌-
❌special_zeta❌-
✅split❌-
✅square✅-
✅stack❌-
✅std❌-
✅std_mean❌-
✅stft❌-
✅t✅-
✅take❌-
✅take_along_dim❌-
✅tanhshrink❌-
✅tensor_split❌-
✅tensordot❌-
✅threshold❌-
✅tile❌-
✅to❌-
❌trace❌-
✅transposeswapaxes, swapdims✅-
✅trapezoid❌-
✅trapz❌-
❌triangular_solve❌-
✅tril❌-
✅tril_indices❌-
✅triplet_margin_loss❌-
✅triu❌-
✅triu_indices❌-
✅type_as❌-
✅unbind❌-
✅unflatten❌-
❌unflatten_dense_tensors❌-
✅unfold❌-
❌uniform✅-
❌unique_consecutive❌-
❌unique_dim❌-
❌unique_dim_consecutive❌-
✅unsafe_chunk❌-
✅unsafe_split❌-
✅unsafe_split_with_sizes❌-
✅upsample_bicubic2d❌-
✅upsample_linear1d❌-
✅upsample_nearest1d❌-
✅upsample_nearest3d❌-
✅upsample_trilinear3d❌-
✅vander❌-
✅var_mean❌-
✅vdot❌-
✅view_as❌-
✅view_as_complex❌-
✅view_as_real❌-
✅vsplit❌-
✅vstackrow_stack❌-
✅xlogyspecial_xlogy❌-
✅zero✅-
✅zeros❌-
✅zeros_like❌-

ONNX support measured by exporting real modules with torch.onnx.export(dynamo=True), one per operator, over 25 generated examples each (hypothesis ci profile). Produced by tox -e proptest_onnx; see how to refresh it.

Total operators exportable, over the same denominators as the TractNNEF tab:

  • core PyTorch opset:

137/138

  • and support from full aten:::

416/555

(headline accounting: 137 core / 416 overall are measured fully exportable here. Of the 555 operators actually measured: 416 full, 2 partial, 125 none, 12 blocked before ONNX, so 137/138 of the measured core operators and 416/555 of all measured ones export fully. Measured with torch 2.13.0, onnxruntime 1.28.0, opset 18.)

How to read this table

The export column grades operator coverage only:
✅ every generated example exported
🟡 some examples exported, others raised
❌ no example exported
⚠️ torch.export could not capture the module, so the ONNX exporter never ran. Not an ONNX verdict

Only ✅ / 🟡 / ❌ are measurements.

runtime (does onnxruntime load and run the exported graph) and numerics (do its outputs match PyTorch) are separate columns on purpose: a graph that exports but diverges numerically is usually a property of the kernel that ran, not a missing operator.

spec coverage says whether this row was measured here, or why it has no direct proptest measurement. Empty categories are not shown as filters. Names that cannot be tied to an attributable operator measurement are listed in the appendix outside both denominators.

The documented column is what the retired TorchScript-exporter listing (this page) said, kept as historical context rather than as a measurement.

exportaten namealiasesruntimenumericsspec coveragedocumentedcan in-placeis core
✅absabsolutefullmatchmeasured✅✅✅
✅acosarccosfullmatchmeasured✅✅✅
✅acosharccoshfullmatchmeasured-✅✅
✅adaptive_avg_pool1dfullmatchmeasured✅❌✅
✅addfullmatchmeasured✅✅✅
✅addmmfullmatchmeasured✅✅✅
✅aliasfullmatchmeasured✅❌✅
✅amaxfullmatchmeasured✅❌✅
✅aminfullmatchmeasured✅❌✅
✅anyfullmatchmeasured✅❌✅
✅arangefullmatchmeasured✅❌✅
✅argmaxfullmatchmeasured✅❌✅
✅argminfullmatchmeasured✅❌✅
✅as_stridedfullmatchmeasured✅✅✅
✅asinarcsinfullmatchmeasured✅✅✅
✅asinharcsinhfullmatchmeasured-✅✅
✅atanarctanfullmatchmeasured✅✅✅
✅atan2arctan2fullpartialmeasured✅✅✅
✅atanharctanhfullmatchmeasured-✅✅
✅avg_pool1dfullmatchmeasured✅❌✅
✅avg_pool2dfullmatchmeasured✅❌✅
✅avg_pool3dfullmatchmeasured✅❌✅
✅bitwise_andfullmatchmeasured✅✅✅
✅bitwise_notfullmatchmeasured✅✅✅
✅bitwise_orfullmatchmeasured✅✅✅
✅bitwise_xorfullmatchmeasured-✅✅
✅bmmfullmatchmeasured✅❌✅
✅catconcat, concatenatefullmatchmeasured✅❌✅
✅ceilfullmatchmeasured✅✅✅
✅clampclipfullmatchmeasured✅✅✅
✅clonefullmatchmeasured✅❌✅
✅col2imfullmatchmeasured✅❌✅
✅constant_pad_ndfullmatchmeasured✅❌✅
✅convolutionfullmatchmeasured✅❌✅
✅copyfullmatchmeasured-✅✅
✅cosfullmatchmeasured✅✅✅
✅coshfullmatchmeasured-✅✅
✅cumsumfullmatchmeasured✅✅✅
✅diagonalfullmatchmeasured✅❌✅
✅divdivide, true_dividefullmatchmeasured✅✅✅
✅elufullmatchmeasured✅✅✅
✅embeddingfullmatchmeasured✅❌✅
✅emptyfull-measured✅❌✅
✅empty_stridedfull-measured-❌✅
✅eqfullmatchmeasured✅✅✅
✅erfspecial_erffullmatchmeasured✅✅✅
✅expfullmatchmeasured✅✅✅
✅expandfullmatchmeasured✅❌✅
✅expm1special_expm1fullmatchmeasured-✅✅
✅fillfullmatchmeasured✅✅✅
✅flipfullmatchmeasured✅❌✅
✅floorfullmatchmeasured✅✅✅
✅fmodfullmatchmeasured✅✅✅
✅fullfullmatchmeasured✅❌✅
✅full_likefullmatchmeasured✅❌✅
✅gatherfullmatchmeasured✅❌✅
✅gegreater_equalfullmatchmeasured✅✅✅
✅gelufullmatchmeasured✅✅✅
✅grid_sampler_2dfullmatchmeasured-❌✅
✅gtgreaterfullmatchmeasured✅✅✅
✅hardtanhfullmatchmeasured✅✅✅
✅indexfullmatchmeasured✅❌✅
✅index_putfullmatchmeasured✅✅✅
✅index_selectfullmatchmeasured✅❌✅
✅isinffullmatchmeasured✅❌✅
✅isnanfullmatchmeasured✅❌✅
✅leless_equalfullmatchmeasured✅✅✅
✅leaky_relufullmatchmeasured✅✅✅
✅logfullmatchmeasured✅✅✅
✅log10fullmatchmeasured✅✅✅
✅log1pspecial_log1pfullmatchmeasured✅✅✅
✅log2fullmatchmeasured✅✅✅
✅logical_andfullmatchmeasured✅✅✅
✅logical_notfullmatchmeasured✅✅✅
✅logical_orfullmatchmeasured✅✅✅
✅logical_xorfullmatchmeasured✅✅✅
✅ltlessfullmatchmeasured✅✅✅
✅masked_scatterfullmatchmeasured✅✅✅
✅maxfullmatchmeasured✅❌✅
✅max_pool2d_with_indicesfullmatchmeasured✅❌✅
✅max_pool3d_with_indicesfullmatchmeasured✅❌✅
✅maximumfullmatchmeasured✅❌✅
✅meanfullmatchmeasured✅❌✅
✅minfullmatchmeasured✅❌✅
✅minimumfullmatchmeasured✅❌✅
✅mmfullmatchmeasured✅❌✅
✅mulmultiplyfullmatchmeasured✅✅✅
✅native_dropoutfullmatchmeasured✅❌✅
✅native_group_normfullpartialmeasured-❌✅
✅native_layer_normfullpartialmeasured✅❌✅
✅nenot_equalfullmatchmeasured✅✅✅
✅negnegativefullmatchmeasured✅✅✅
✅nonzerofullmatchmeasured✅❌✅
✅permutefullmatchmeasured✅❌✅
✅powfullmatchmeasured✅✅✅
✅prodfullmatchmeasured✅❌✅
✅randfull-measured✅❌✅
✅randnfull-measured✅❌✅
❌randperm--measured-❌✅
✅reciprocalfullmatchmeasured✅✅✅
✅reflection_pad1dfullmatchmeasured✅❌✅
✅reflection_pad2dfullmatchmeasured✅❌✅
✅reflection_pad3dfullmatchmeasured✅❌✅
✅relufullmatchmeasured✅✅✅
✅remainderfullpartialmeasured✅✅✅
✅repeatfullmatchmeasured✅❌✅
✅replication_pad2dfullmatchmeasured✅❌✅
✅replication_pad3dfullmatchmeasured✅❌✅
✅roundspecial_roundfullmatchmeasured✅✅✅
✅rsqrtfullmatchmeasured✅✅✅
✅scalar_tensorfullmatchmeasured✅❌✅
✅scatterfullmatchmeasured✅✅✅
✅scatter_addfullmatchmeasured✅✅✅
✅scatter_reducefullmatchmeasured✅✅✅
✅selectfullmatchmeasured✅❌✅
✅select_scatterfullmatchmeasured-❌✅
✅sigmoidspecial_expitfullmatchmeasured✅✅✅
✅signfullmatchmeasured✅✅✅
✅sinfullmatchmeasured✅✅✅
✅sinhfullmatchmeasured-✅✅
✅slicefullmatchmeasured✅❌✅
✅slice_scatterfullmatchmeasured-❌✅
✅sortfullmatchmeasured✅❌✅
✅split_with_sizesfullmatchmeasured✅❌✅
✅sqrtfullmatchmeasured✅✅✅
✅squeezefullmatchmeasured✅✅✅
✅subsubtractfullmatchmeasured✅✅✅
✅sumfullpartialmeasured✅❌✅
✅tanfullmatchmeasured✅✅✅
✅tanhfullmatchmeasured✅✅✅
✅topkfullmatchmeasured✅❌✅
✅truncfixfullmatchmeasured-✅✅
✅unsqueezefullmatchmeasured✅✅✅
✅upsample_bilinear2dfullmatchmeasured✅❌✅
✅upsample_nearest2dfullmatchmeasured✅❌✅
✅varfullmatchmeasured✅❌✅
✅viewfullmatchmeasured✅❌✅
✅wherefullmatchmeasured✅❌✅
✅adaptive_avg_pool2dfullmatchmeasured✅❌-
❌adaptive_avg_pool3d--measured✅❌-
✅adaptive_max_pool1dfullmatchmeasured✅❌-
✅adaptive_max_pool2dfullmatchmeasured✅❌-
✅adaptive_max_pool3dfullmatchmeasured✅❌-
✅addbmmfullmatchmeasured-✅-
✅addcdivfullmatchmeasured-❌-
✅addcmulfullmatchmeasured✅❌-
✅addmvfullmatchmeasured-✅-
✅addrfullmatchmeasured-✅-
✅affine_grid_generatorfullmatchmeasured✅❌-
✅allfullmatchmeasured✅❌-
✅alpha_dropoutfullmatchmeasured✅✅-
✅aminmaxfullmatchmeasured✅❌-
✅anglefullmatchmeasured-❌-
✅argsortfullmatchmeasured✅❌-
✅argwherefullmatchmeasured-❌-
✅atleast_1dfullmatchmeasured✅❌-
✅atleast_2dfullmatchmeasured✅❌-
✅atleast_3dfullmatchmeasured✅❌-
✅baddbmmfullmatchmeasured✅✅-
❌bartlett_window--measured-❌-
✅batch_normfullmatchmeasured✅❌-
✅bernoullifull-measured✅✅-
✅bilinearfullmatchmeasured-❌-
✅binary_cross_entropyfullmatchmeasured-❌-
✅binary_cross_entropy_with_logitsfullmatchmeasured✅❌-
❌bincount--measured-❌-
❌binomial--measured-❌-
✅bitwise_left_shiftfullmatchmeasured✅✅-
✅bitwise_right_shiftfullmatchmeasured✅✅-
✅blackman_windowfullmatchmeasured-❌-
✅block_diagfullmatchmeasured-❌-
✅broadcast_tensorsfullmatchmeasured✅❌-
✅broadcast_tofullmatchmeasured✅❌-
✅bucketizefullmatchmeasured✅❌-
✅cartesian_prodfullmatchmeasured-❌-
✅cauchyfull-measured-✅-
❌cdist--measured✅❌-
✅celufullmatchmeasured✅❌-
✅chain_matmulfullmatchmeasured-❌-
❌chalf--measured-❌-
✅channel_shufflefullmatchmeasured-❌-
❌cholesky--measured-❌-
❌cholesky_inverse--measured-❌-
❌cholesky_solve--measured-❌-
✅chunkfullmatchmeasured✅❌-
✅clamp_maxfullmatchmeasured✅✅-
✅clamp_minfullmatchmeasured✅✅-
✅column_stackfullmatchmeasured-❌-
⚠️combinations--measured-❌-
✅complexfullmatchmeasured-❌-
✅conjfullmatchmeasured-❌-
🟡conj_physicalfullmatchmeasured✅✅-
✅contiguousfullmatchmeasured✅❌-
✅conv1dfullmatchmeasured✅❌-
✅conv2dfullmatchmeasured✅❌-
✅conv3dfullmatchmeasured✅❌-
❌conv_tbc--measured✅❌-
✅conv_transpose1dfullmatchmeasured✅❌-
✅conv_transpose2dfullmatchmeasured✅❌-
✅conv_transpose3dfullmatchmeasured✅❌-
❌copy_sparse_to_sparse_--measured-❌-
❌copysign--measured-✅-
⚠️corrcoef--measured-❌-
✅cosine_embedding_lossfullmatchmeasured-❌-
✅cosine_similarityfullmatchmeasured✅❌-
✅count_nonzerofullmatchmeasured-❌-
✅crossfullmatchmeasured✅❌-
✅cross_entropy_lossfullmatchmeasured✅❌-
⚠️ctc_loss--measured-❌-
❌cummax--measured-❌-
❌cummin--measured-❌-
❌cumprod--measured-✅-
✅cumulative_trapezoidfullmatchmeasured-❌-
✅deg2radfullmatchmeasured-❌-
❌dequantize--measured✅❌-
✅detachfullmatchmeasured✅✅-
✅diagfullmatchmeasured-❌-
✅diag_embedfullmatchmeasured-❌-
✅diagflatfullmatchmeasured-❌-
✅difffullmatchmeasured-❌-
❌digammaspecial_digamma, special_psi--measured-✅-
✅distfullmatchmeasured-❌-
✅dotfullmatchmeasured✅❌-
✅dropoutfullmatchmeasured✅✅-
✅dsplitfullmatchmeasured-❌-
✅dstackfullmatchmeasured-❌-
✅einsumfullmatchmeasured✅❌-
✅embedding_bagfullmatchmeasured✅❌-
✅embedding_renorm_fullmatchmeasured-❌-
✅empty_likefullmatchmeasured✅❌-
❌empty_permuted--measured-❌-
⚠️empty_quantized--measured-❌-
✅erfcspecial_erfcfullmatchmeasured-✅-
❌erfinvspecial_erfinv--measured-✅-
✅exp2special_exp2fullmatchmeasured-✅-
✅expand_asfullmatchmeasured✅❌-
✅exponentialfull-measured-✅-
✅eyefullmatchmeasured✅❌-
✅feature_alpha_dropoutfullmatchmeasured✅✅-
✅feature_dropoutfullmatchmeasured✅✅-
❌fft_fftfreq--measured-❌-
❌fft_ihfft2--measured-❌-
❌fft_ihfftn--measured-❌-
✅fft_irfftnfullmatchmeasured-❌-
❌fft_rfftfreq--measured-❌-
✅fft_rfftnfullpartialmeasured-❌-
✅flattenfullmatchmeasured✅❌-
✅flatten_dense_tensorsfullmatchmeasured-❌-
✅fliplrfullmatchmeasured-❌-
✅flipudfullmatchmeasured-❌-
✅float_powerfullmatchmeasured-✅-
✅floor_dividefullmatchmeasured✅✅-
❌fmax--measured-❌-
❌fmin--measured-❌-
✅fracfullmatchmeasured-❌-
❌fractional_max_pool2d--measured-❌-
❌fractional_max_pool3d--measured-❌-
❌frexp--measured-❌-
✅frobenius_normfullmatchmeasured✅❌-
❌gamma--measured-❌-
❌gcd--measured-✅-
✅geometricfull-measured-✅-
❌geqrf--measured-❌-
✅glufullmatchmeasured✅❌-
✅grid_samplerfullpartialmeasured✅❌-
❌grid_sampler_3d--measured-❌-
✅group_normfullmatchmeasured✅❌-
✅grufullmatchmeasured✅❌-
✅hamming_windowfulldivergemeasured-❌-
✅hann_windowfullmatchmeasured✅❌-
✅hardshrinkfullmatchmeasured✅✅-
✅hardsigmoidfullmatchmeasured✅✅-
✅hardswishfullmatchmeasured✅✅-
✅heavisidefullmatchmeasured-❌-
✅hinge_embedding_lossfullmatchmeasured-❌-
❌histc--measured-❌-
❌histogram--measured-❌-
❌histogramdd--measured-❌-
✅hsplitfullmatchmeasured-❌-
✅hstackfullmatchmeasured✅❌-
✅huber_lossfullmatchmeasured-❌-
❌hypot--measured-✅-
❌i0special_i0--measured-✅-
❌igammaspecial_gammainc--measured-✅-
❌igammacspecial_gammaincc--measured-✅-
✅im2colfullmatchmeasured✅❌-
✅imagfullmatchmeasured-❌-
✅index_addfullmatchmeasured✅❌-
✅index_copyfullmatchmeasured✅❌-
✅index_fillfullmatchmeasured✅❌-
❌index_put_impl_--measured-❌-
❌index_reduce--measured-✅-
✅innerfullmatchmeasured-❌-
✅instance_normfullpartialmeasured✅❌-
✅isclosefullpartialmeasured-❌-
✅isfinitefullmatchmeasured✅❌-
✅isinfullmatchmeasured-❌-
✅isneginffullmatchmeasured-❌-
✅isposinffullmatchmeasured-❌-
✅isrealfullmatchmeasured-❌-
❌istft--measured-❌-
❌kaiser_window--measured-❌-
✅kl_divfullmatchmeasured✅❌-
✅kronfullmatchmeasured-❌-
❌kthvalue--measured-❌-
✅l1_lossfullmatchmeasured-❌-
✅layer_normfullpartialmeasured✅❌-
❌lcm--measured-✅-
✅ldexpfullmatchmeasured-✅-
✅lerpfullmatchmeasured✅✅-
❌lgamma--measured-✅-
❌linalg_cond--measured-❌-
✅linalg_crossfullmatchmeasured✅❌-
✅linalg_detdetfullmatchmeasured✅❌-
✅linalg_diagonalfullmatchmeasured-❌-
❌linalg_eig--measured-❌-
❌linalg_eigh--measured-❌-
❌linalg_eigvals--measured-❌-
❌linalg_eigvalsh--measured-❌-
❌linalg_householder_productorgqr--measured-❌-
❌linalg_invinverse--measured-❌-
❌linalg_ldl_solve--measured-❌-
⚠️linalg_lstsq--measured-❌-
❌linalg_lu--measured-❌-
❌linalg_lu_solve--measured-❌-
❌linalg_matrix_expmatrix_exp--measured-❌-
✅linalg_matrix_normfullmatchmeasured✅❌-
✅linalg_matrix_powermatrix_powerfullmatchmeasured-❌-
❌linalg_matrix_rank--measured-❌-
✅linalg_normfullmatchmeasured✅❌-
❌linalg_pinv--measured-❌-
❌linalg_qr--measured-❌-
❌linalg_slogdet--measured-❌-
❌linalg_solve--measured-❌-
❌linalg_solve_triangular--measured-❌-
❌linalg_svd--measured-❌-
❌linalg_tensorinv--measured-❌-
❌linalg_tensorsolve--measured-❌-
✅linalg_vector_normfullmatchmeasured✅❌-
✅linearfullmatchmeasured✅❌-
✅linspacefullmatchmeasured✅❌-
✅log_normalfull-measured-✅-
✅log_sigmoidfullmatchmeasured✅❌-
✅log_softmaxspecial_log_softmaxfullmatchmeasured✅❌-
✅logaddexpfullmatchmeasured-❌-
✅logaddexp2fullmatchmeasured-❌-
✅logcumsumexpfullmatchmeasured-❌-
✅logdetfullmatchmeasured✅❌-
✅logitspecial_logitfullmatchmeasured✅✅-
✅logspacefullmatchmeasured-❌-
✅logsumexpspecial_logsumexpfullmatchmeasured✅❌-
✅lstmfullmatchmeasured✅❌-
❌lu_solve--measured-❌-
❌lu_unpack--measured-❌-
✅mHadjointfullmatchmeasured-❌-
✅mTfullmatchmeasured-❌-
✅margin_ranking_lossfullmatchmeasured-❌-
✅masked_fillfullmatchmeasured✅✅-
❌masked_select--measured✅❌-
✅matmullinalg_matmulfullmatchmeasured✅❌-
✅matrix_Hfullmatchmeasured-❌-
✅max_pool1dfullmatchmeasured✅❌-
✅max_pool1d_with_indicesfullmatchmeasured✅❌-
✅max_pool2dfullmatchmeasured✅❌-
✅max_pool3dfullmatchmeasured✅❌-
✅max_unpool2dfullmatchmeasured-❌-
✅max_unpool3dfullmatchmeasured-❌-
❌median--measured-❌-
✅meshgridfullmatchmeasured✅❌-
✅mishfullmatchmeasured✅❌-
❌mode--measured-❌-
✅movedimmoveaxisfullmatchmeasured✅❌-
✅mse_lossfullmatchmeasured✅❌-
✅msortfullmatchmeasured-❌-
✅multi_margin_lossfullmatchmeasured-❌-
✅multilabel_margin_lossfullmatchmeasured-❌-
✅multinomialfull-measured✅❌-
✅mvfullmatchmeasured✅❌-
❌mvlgammaspecial_multigammaln--measured-❌-
✅nan_to_numfullmatchmeasured✅✅-
✅nanmeanfullmatchmeasured-❌-
❌nanmedian--measured-❌-
✅nanquantilefullmatchmeasured-❌-
✅nansumfullmatchmeasured-❌-
✅narrowfullmatchmeasured✅❌-
❌native_channel_shuffle--measured-❌-
✅new_emptyfullmatchmeasured✅❌-
✅new_empty_stridedfull-measured-❌-
✅new_fullfullmatchmeasured✅❌-
✅new_onesfullmatchmeasured✅❌-
✅new_zerosfullmatchmeasured✅❌-
❌nextafter--measured-✅-
✅nll_lossfullmatchmeasured✅❌-
✅nll_loss2dfullmatchmeasured✅❌-
✅nll_loss_ndfullmatchmeasured✅❌-
✅nonzero_numpyfullmatchmeasured✅❌-
❌nonzero_static--measured-❌-
✅normfullmatchmeasured✅❌-
✅normalfull-measured✅✅-
❌nuclear_norm--measured-❌-
✅numpy_Tfullmatchmeasured✅❌-
✅one_hotfullmatchmeasured✅❌-
✅onesfullmatchmeasured✅❌-
✅ones_likefullmatchmeasured✅❌-
❌ormqr--measured-❌-
✅outergerfullmatchmeasured✅❌-
✅padfullmatchmeasured✅❌-
✅pad_sequencefullmatchmeasured-❌-
✅pairwise_distancefullmatchmeasured✅❌-
❌pdist--measured-❌-
✅pixel_shufflefullmatchmeasured✅❌-
✅pixel_unshufflefullmatchmeasured✅❌-
❌poisson--measured-❌-
✅poisson_nll_lossfullmatchmeasured-❌-
✅polarfullmatchmeasured-❌-
❌polygammaspecial_polygamma--measured-✅-
✅positivefullmatchmeasured-❌-
✅prelufullmatchmeasured✅❌-
❌put--measured-✅-
❌qr--measured-❌-
✅quantilefullmatchmeasured-❌-
⚠️quantize_per_channel--measured-❌-
❌quantize_per_tensor--measured✅❌-
⚠️quantize_per_tensor_dynamic--measured-❌-
⚠️quantized_batch_norm--measured-❌-
⚠️quantized_max_pool1d--measured-❌-
⚠️quantized_max_pool2d--measured-❌-
⚠️quantized_max_pool3d--measured-❌-
✅rad2degfullmatchmeasured-❌-
✅rand_likefull-measured✅❌-
✅randintfull-measured✅❌-
✅randint_likefull-measured✅❌-
✅randn_likefull-measured✅❌-
⚠️random--measured-✅-
❌range--measured-❌-
✅ravelfullmatchmeasured-❌-
✅realfullmatchmeasured-❌-
✅relu6fullmatchmeasured✅✅-
✅renormfullmatchmeasured-❌-
✅repeat_interleavefullpartialmeasured✅❌-
✅replication_pad1dfullmatchmeasured✅❌-
✅reshapefullmatchmeasured✅❌-
✅reshape_asfullmatchmeasured✅❌-
❌resize_as_sparse_--measured-❌-
✅resolve_conjfullmatchmeasured✅❌-
✅resolve_negfullmatchmeasured✅❌-
✅rnn_relufullmatchmeasured✅❌-
✅rnn_tanhfullmatchmeasured✅❌-
✅rollfullmatchmeasured✅❌-
✅rot90fullmatchmeasured-❌-
❌rrelu--measured✅❌-
❌rrelu_with_noise--measured-✅-
✅rsubfullmatchmeasured✅❌-
✅scaled_dot_product_attentionfullmatchmeasured✅❌-
❌searchsorted--measured-❌-
❌segment_reduce--measured-❌-
✅selufullmatchmeasured✅✅-
✅set_fullmatchmeasured-❌-
❌sgn--measured-❌-
✅signbitfullmatchmeasured-❌-
✅silufullmatchmeasured✅✅-
✅sincspecial_sincfullmatchmeasured-❌-
✅smooth_l1_lossfullmatchmeasured-❌-
✅soft_margin_lossfullmatchmeasured-❌-
✅softmaxspecial_softmaxfullmatchmeasured✅❌-
✅softplusfullmatchmeasured✅❌-
✅softshrinkfullmatchmeasured✅❌-
❌special_airy_ai--measured-❌-
❌special_bessel_j0--measured-❌-
❌special_bessel_j1--measured-❌-
❌special_bessel_y0--measured-❌-
❌special_bessel_y1--measured-❌-
❌special_chebyshev_polynomial_t--measured-❌-
❌special_chebyshev_polynomial_u--measured-❌-
❌special_chebyshev_polynomial_v--measured-❌-
❌special_chebyshev_polynomial_w--measured-❌-
✅special_entrfullmatchmeasured-❌-
✅special_erfcxfullpartialmeasured-❌-
❌special_hermite_polynomial_h--measured-❌-
❌special_hermite_polynomial_he--measured-❌-
❌special_i0e--measured-❌-
❌special_i1--measured-❌-
❌special_i1e--measured-❌-
❌special_laguerre_polynomial_l--measured-❌-
❌special_legendre_polynomial_p--measured-❌-
✅special_log_ndtrfullpartialmeasured-❌-
❌special_modified_bessel_i0--measured-❌-
❌special_modified_bessel_i1--measured-❌-
❌special_modified_bessel_k0--measured-❌-
❌special_modified_bessel_k1--measured-❌-
✅special_ndtrfullmatchmeasured-❌-
❌special_ndtri--measured-❌-
❌special_scaled_modified_bessel_k0--measured-❌-
❌special_scaled_modified_bessel_k1--measured-❌-
❌special_shifted_chebyshev_polynomial_t--measured-❌-
❌special_shifted_chebyshev_polynomial_u--measured-❌-
❌special_shifted_chebyshev_polynomial_v--measured-❌-
❌special_shifted_chebyshev_polynomial_w--measured-❌-
❌special_spherical_bessel_j0--measured-❌-
✅special_xlog1pyfullmatchmeasured-❌-
❌special_zeta--measured-❌-
✅splitfullpartialmeasured✅❌-
✅squarefullmatchmeasured✅✅-
✅stackfullmatchmeasured✅❌-
✅stdfullpartialmeasured✅❌-
✅std_meanfullpartialmeasured✅❌-
✅stftfullmatchmeasured✅❌-
✅tfullmatchmeasured✅✅-
✅takefullmatchmeasured✅❌-
✅take_along_dimfullmatchmeasured-❌-
✅tanhshrinkfullmatchmeasured-❌-
✅tensor_splitfullmatchmeasured✅❌-
✅tensordotfullmatchmeasured✅❌-
✅thresholdfullmatchmeasured✅❌-
✅tilefullmatchmeasured✅❌-
✅tofullmatchmeasured✅❌-
✅tracefullmatchmeasured-❌-
✅transposeswapaxes, swapdimsfullmatchmeasured✅✅-
✅trapezoidfullmatchmeasured-❌-
✅trapzfullmatchmeasured-❌-
❌triangular_solve--measured-❌-
✅trilfullmatchmeasured✅❌-
✅tril_indicesfullmatchmeasured-❌-
✅triplet_margin_lossfullmatchmeasured-❌-
✅triufullmatchmeasured✅❌-
✅triu_indicesfullmatchmeasured-❌-
✅type_asfullmatchmeasured✅❌-
✅unbindfullmatchmeasured✅❌-
✅unflattenfullmatchmeasured✅❌-
✅unflatten_dense_tensorsfullmatchmeasured-❌-
✅unfoldfullmatchmeasured✅❌-
❌uniform--measured-✅-
🟡unique_consecutivefullmatchmeasured-❌-
✅unique_dimfullmatchmeasured✅❌-
❌unique_dim_consecutive--measured-❌-
✅unsafe_chunkfullmatchmeasured✅❌-
✅unsafe_splitfullmatchmeasured✅❌-
✅unsafe_split_with_sizesfullmatchmeasured✅❌-
✅upsample_bicubic2dfullpartialmeasured✅❌-
✅upsample_linear1dfullmatchmeasured✅❌-
✅upsample_nearest1dfullmatchmeasured✅❌-
✅upsample_nearest3dfullmatchmeasured✅❌-
✅upsample_trilinear3dfullpartialmeasured✅❌-
❌vander--measured-❌-
✅var_meanfullmatchmeasured✅❌-
✅vdotfullmatchmeasured-❌-
✅view_asfullmatchmeasured✅❌-
✅view_as_complexfullmatchmeasured-❌-
✅view_as_realfullmatchmeasured-❌-
✅vsplitfullmatchmeasured-❌-
✅vstackrow_stackfullmatchmeasured✅❌-
✅xlogyspecial_xlogyfullmatchmeasured-❌-
✅zerofullmatchmeasured✅✅-
✅zerosfullmatchmeasured✅❌-
✅zeros_likefullmatchmeasured✅❌-

Appendix: identifiers excluded from this page

These names are filtered out of the support tables above because the generator cannot turn them into an attributable inference operator measurement. Each group has a documented rationale; click any to expand the full member list.

Python builtin / scripting false positives (52 names)

capitalize, center, chr, clear, dict, endswith, expandtabs, extend, find, format, get, getelem, hash, hex, isalnum, isalpha, isdecimal, isdigit, isidentifier, islower, isnumeric, isprintable, isspace, istitle, isupper, items, join, keys, ljust, lower, lstrip, oct, ord, partition, popitem, rfind, rindex, rjust, rpartition, rsplit, rstrip, setdefault, sorted, splitlines, startswith, strip, swapcase, title, update, upper, values, zfill

Test harness / placeholder identifiers (13 names)

confirmed_by_owner, foo, mathremainder, percentFormat, pointwise_placeholder, symbolic_b, test, test_symbol, test_vartype, test_vartype2, unknown, view_expand_placeholder, your_op

Distributed / RPC primitives (14 names)

all_gather_into_tensor, all_reduce, fork, get_gradients, is_owner, local_value, owner, owner_name, reduce_scatter_tensor, to_here, wait, wait_tensor, warn, warns

Backend-specific dispatch shims (cudnn / miopen / mkldnn / mps) (26 names)

cudnn_affine_grid_generator, cudnn_batch_norm, cudnn_convolution, cudnn_convolution_add_relu, cudnn_convolution_relu, cudnn_convolution_transpose, cudnn_grid_sampler, cudnn_is_acceptable, miopen_batch_norm, miopen_convolution, miopen_convolution_add_relu, miopen_convolution_relu, miopen_convolution_transpose, miopen_ctc_loss, miopen_depthwise_convolution, miopen_rnn, mkldnn_adaptive_avg_pool2d, mkldnn_convolution, mkldnn_linear, mkldnn_max_pool2d, mkldnn_max_pool3d, mkldnn_reorder_conv2d_weight, mkldnn_reorder_conv3d_weight, mkldnn_rnn_layer, mps_linear, to_mkldnn

Tensor metadata accessors (return a Python value, not a tensor) (60 names)

can_cast, data, dense_dim, device, dtype, element_size, enable_grad, get_autocast_dtype, get_device, get_pool_ceil_padding, grad, has_torch_function, iinfo, initial_seed, int_repr, is_autocast_cpu_enabled, is_autocast_enabled, is_coalesced, is_complex, is_conj, is_contiguous, is_cuda, is_grad_enabled, is_leaf, is_non_overlapping_and_dense, is_nonzero, is_pinned, is_same_size, is_scripting, is_set_to, is_signed, is_strides_like_format, manual_seed, node, op, op_name, output_nr, pin_memory, promote_types, q_per_channel_axis, q_per_channel_scales, q_per_channel_zero_points, q_scale, q_zero_point, qscheme, record_stream, refine_names, rename, requires_grad_, result_type, retain_grad, retains_grad, save, seed, set_data, set_grad_enabled, set_source_Tensor_storage_offset, sparse_dim, storage_offset, stride

Named-tensor API (names stripped before JIT trace) (3 names)

align_as, align_tensors, align_to

Sparse-tensor machinery (NNEF / tract are dense-only) (25 names)

ccol_indices, ccol_indices_copy, coalesce, col_indices, col_indices_copy, copy_sparse_to_sparse, crow_indices, crow_indices_copy, hspmm, indices, indices_copy, nested_to_padded_tensor, row_indices, row_indices_copy, smm, sparse_compressed_tensor, sparse_coo_tensor, sparse_mask, sparse_resize, sparse_resize_and_clear, sparse_sampled_addmm, sspaddmm, to_dense, to_padded_tensor, values_copy

Functionalization *_copy / *_scatter variants (20 names)

alias_copy, as_strided_copy, as_strided_scatter, detach_copy, diagonal_copy, diagonal_scatter, lift, lift_fresh, lift_fresh_copy, narrow_copy, permute_copy, select_copy, slice_copy, slice_inverse, split_copy, unbind_copy, unfold_copy, view_as_complex_copy, view_as_real_copy, view_copy

Batch-norm training / backward-only internals (8 names)

batch_norm_elemt, batch_norm_gather_stats, batch_norm_gather_stats_with_counts, batch_norm_stats, batch_norm_update_stats, native_batch_norm, native_norm, norm_except_dim

linalg_*_ex paired-output variants + deprecated linalg wrappers (12 names)

eig, linalg_cholesky_ex, linalg_inv_ex, linalg_ldl_factor_ex, linalg_lu_factor_ex, lstsq, matrix_rank, pinv, pinverse, solve, svd, symeig

QAT fake_quantize_* training-only ops (6 names)

choose_qparams_optimized, fake_quantize_per_channel_affine, fake_quantize_per_channel_affine_cachemask, fake_quantize_per_tensor_affine, fake_quantize_per_tensor_affine_cachemask, fused_moving_avg_obs_fake_quant

slow_conv* / thnn_conv* dispatcher-fallback kernels (8 names)

conv_depthwise3d, slow_conv3d, slow_conv3d_forward, slow_conv_dilated2d, slow_conv_dilated3d, slow_conv_transpose2d, slow_conv_transpose3d, thnn_conv2d

Python / TorchScript scalar builtins (extra) (26 names)

append, bin, count, cpu, cuda, degrees, dim, divmod, equal, fabs, factorial, insert, is_floating_point, item, len, list, list_with_default, modf, neq, pop, radians, remove, replace, reverse, str, tensor

Proptest rows with no attributable graph target (26 names)

allclose, as_tensor, bias_addmm, broadcast_shapes, convolution_overrideable, convrelu, copy_to, first, floordiv, gru_cell, hash_tensor, linalg__powsum, lstm_cell, native_multi_head_self_attention, numel, quantize, quantized_gru, quantized_lstm, resize_, rnn_relu_cell, rnn_tanh_cell, second, size, upsample, wrapped_linear_prepack, wrapped_quantized_linear_prepacked

Inplace storage / metadata mutators stripped by JIT (9 names)

fill_diagonal_, float_power_, rename_, resize, resize_as_, resize_as_sparse, set, sparse_resize_, sparse_resize_and_clear_

Backward / dynamo-autograd internals (10 names)

embedding_renorm, from_file, glu_jvp, log_sigmoid_forward, multilabel_margin_loss_forward, nll_loss_forward, normal_functional, rowwise_prune, rrelu_with_noise_functional, sum_to

Regex-scrape artifacts (phantom names from f-strings) (1 names)

conv