onnx/onnx

12.3

Adjusted Score

12.3

Raw Score

100%

Time Factor

2026-08-28

Last Push

21.4K

Stars

Python

Language

173.9K

Lines of Code

837

Files

2.2K

Pattern Hits

2026-08-29

Scan Date

0.06

HC Hit Rate

What These Metrics Mean

Adjusted Score: Primary synthetic code indicator. Raw score normalised per 1,000 lines of code and multiplied by the temporal discount factor. This is the definitive comparative metric — use it to rank repositories by AI authorship density.
Raw Score: The unmodified sum of all severity-weighted, context-multiplied pattern match scores before temporal discounting. Reflects the absolute signal strength independent of when the repository was last active.
Time Factor: The temporal discount multiplier (0–100%) applied to the raw score. Repositories last updated before ChatGPT's launch (Nov 2022) receive a 5% factor. Full signal is only assigned to repositories active in the post-adoption era (Jan 2024+).
Pattern Hits: Total count of individual pattern matches across all files and categories. A high hit count with a low score may indicate a very large codebase with isolated AI snippets; a low count with a high score indicates dense, concentrated AI signatures.
HC Hit Rate: High+Critical pattern hits per file, averaged across the repository. This orthogonal signal catches repositories where a few files are densely packed with high-severity AI tells — a strong indicator even when the normalised score appears moderate due to codebase size.
Lines of Code / Files: Total lines and files analysed. The scanner examines 94 file extensions. These denominators are used to normalise the score, enabling fair comparison between repositories of vastly different sizes.

Score History

This chart maps the temporal evolution of the adjusted synthetic code score across successive scan runs. An upward trajectory indicates ongoing incorporation of AI-generated code or expanding LLM-assisted scaffolding; a stable or declining trajectory may reflect active human refactoring, code removal, or the adoption of stricter authorship policies. The dashed secondary line (right axis) independently tracks total raw pattern hit count, which can diverge from the normalised score when codebase size changes significantly between scans.

Severity Breakdown

Classifies detected patterns by their diagnostic confidence and structural impact. CRITICAL patterns (coefficient 10) represent definitive synthetic signatures — hallucinated imports, explicit LLM attribution metadata — virtually never produced by human authors. HIGH (5) indicates strong structural tells such as cross-file repetition or cross-linguistic idioms. MEDIUM (2) covers recognisable conversational padding and AI-specific vocabulary. LOW (1) captures subtle indicators like tautological comments and generic boilerplate that require density to carry independent signal.

CRITICAL 0HIGH 52MEDIUM 43LOW 2144

Directory Score Breakdown

This horizontal bar chart decomposes the repository's raw synthetic code score by top-level directory, allowing you to pinpoint precisely which modules or components carry the highest AI authorship density. Directories with disproportionately high scores relative to their size warrant targeted manual review: concentrated AI signatures often trace back to mass-generated configuration layers, auto-ported test suites, LLM-scaffolded boilerplate classes, or entire subsystems authored under heavy copilot assistance. Use this view to prioritise your human code-review effort.

Pattern Findings

The scanner identified 2239 distinct pattern matches across 20 syntactic categories. Each entry below represents a discrete location in the source code where the engine recorded a statistically significant AI authorship indicator. Expand any category row to inspect the individual file paths, line numbers, code snippets, and the lexical context (CODE, COMMENT, or STRING) in which each match was detected.

Reading the findings table: The Severity column indicates the diagnostic confidence level (CRITICAL / HIGH / MEDIUM / LOW). The Context column identifies whether the match occurred inside executable code, an inline comment, or a string literal — comment-context matches receive a ×1.5 weight because LLMs systematically over-annotate. The ⚡ bolt icon marks clustered matches: three or more patterns within a 10-line window, each receiving an additional ×1.5 density multiplier as dense clusters constitute far stronger evidence of synthetic authorship than isolated hits.

Hyper-Verbose Identifiers1008 hits · 749 pts

Severity	File	Line	Snippet	Context
LOW	onnx/model_container.py	120	def is_in_memory_external_initializer(self, name: str) -> bool:	CODE
LOW	onnx/__init__.py	462	def _operator_set_protos_repr(protos: Sequence[OperatorSetIdProto]) -> str:	CODE
LOW	onnx/shape_inference.py	159	def infer_function_output_types(	CODE
LOW	onnx/utils.py	66	def _dfs_search_reachable_nodes(	CODE
LOW	onnx/utils.py	109	def _collect_referred_local_functions(	CODE
LOW	onnx/utils.py	132	def _collect_reachable_tensors(	CODE
LOW	onnx/helper.py	86	def _create_op_set_id_version_map(table: VersionTableType) -> VersionMapType:	CODE
LOW	onnx/helper.py	781	def make_empty_tensor_value_info(name: str) -> ValueInfoProto:	CODE
LOW	onnx/helper.py	850	def make_sparse_tensor_type_proto(	CODE
LOW	onnx/helper.py	895	def make_sparse_tensor_value_info(	CODE
LOW	onnx/helper.py	973	def make_tensor_sequence_value_info(	CODE
LOW	onnx/helper.py	1285	def tensor_dtype_to_storage_tensor_dtype(tensor_dtype: int) -> int:	CODE
LOW	onnx/inliner.py	30	def inline_selected_functions(	CODE
LOW	onnx/external_data_helper.py	112	def _validate_external_data_file_bounds(	CODE
LOW	onnx/external_data_helper.py	147	def load_external_data_for_tensor(tensor: TensorProto, base_dir: str) -> None:	CODE
LOW	onnx/external_data_helper.py	163	def load_external_data_for_model(model: ModelProto, base_dir: str) -> None:	CODE
LOW	onnx/external_data_helper.py	207	def convert_model_to_external_data(	CODE
LOW	onnx/external_data_helper.py	265	def convert_model_from_external_data(model: ModelProto) -> None:	CODE
LOW	onnx/external_data_helper.py	331	def _recursive_attribute_processor(	CODE
LOW	onnx/external_data_helper.py	342	def _get_initializer_tensors_from_graph(graph: GraphProto, /) -> Iterable[TensorProto]:	CODE
LOW	onnx/external_data_helper.py	357	def _get_attribute_tensors_from_graph(	CODE
LOW	onnx/external_data_helper.py	393	def remove_external_data_field(tensor: TensorProto, field_key: str) -> None:	CODE
LOW	onnx/external_data_helper.py	407	def write_external_data_tensors(model: ModelProto, filepath: str) -> ModelProto:	CODE
LOW	onnx/serialization.py	82	def get_format_from_file_extension(self, file_extension: str) -> str \| None:	CODE
LOW	onnx/tools/replace_constants.py	71	def _replace_constant_of_shape_with_range(	CODE
LOW	onnx/tools/replace_constants.py	153	def _replace_constant_of_shape_value(	CODE
LOW	onnx/tools/replace_constants.py	199	def replace_initializer_by_constant_of_shape( # noqa: PLR0911	CODE
LOW	onnx/tools/update_model_dims.py	12	def update_inputs_outputs_dims(	CODE
LOW	onnx/backend/test/stat_coverage.py	256	def gen_overall_test_coverage(	CODE
LOW	onnx/backend/test/case/model/gradient.py	16	def export_gradient_scalar_add() -> None:	CODE
LOW	onnx/backend/test/case/model/gradient.py	61	def export_gradient_scalar_add_and_mul() -> None:	CODE
LOW	onnx/backend/test/case/node/globalmaxpool.py	26	def export_globalmaxpool_precomputed() -> None:	CODE
LOW	…x/backend/test/case/node/softmax_cross_entropy_loss.py	101	def export_softmaxcrossentropy_none() -> None:	CODE
LOW	…x/backend/test/case/node/softmax_cross_entropy_loss.py	125	def export_softmaxcrossentropy_none_log_prob() -> None:	CODE
LOW	…x/backend/test/case/node/softmax_cross_entropy_loss.py	156	def export_softmaxcrossentropy_none_weights() -> None:	CODE
LOW	…x/backend/test/case/node/softmax_cross_entropy_loss.py	186	def export_softmaxcrossentropy_none_weights_log_prob() -> None:	CODE
LOW	…x/backend/test/case/node/softmax_cross_entropy_loss.py	218	def export_softmaxcrossentropy_sum() -> None:	CODE
LOW	…x/backend/test/case/node/softmax_cross_entropy_loss.py	242	def export_softmaxcrossentropy_sum_log_prob() -> None:	CODE
LOW	…x/backend/test/case/node/softmax_cross_entropy_loss.py	273	def export_softmaxcrossentropy_mean() -> None:	CODE
LOW	…x/backend/test/case/node/softmax_cross_entropy_loss.py	297	def export_softmaxcrossentropy_mean_log_prob() -> None:	CODE
LOW	…x/backend/test/case/node/softmax_cross_entropy_loss.py	379	def export_softmaxcrossentropy_mean_weights() -> None:	CODE
LOW	…x/backend/test/case/node/softmax_cross_entropy_loss.py	409	def export_softmaxcrossentropy_mean_weights_log_prob() -> None:	CODE
LOW	…x/backend/test/case/node/softmax_cross_entropy_loss.py	441	def export_softmaxcrossentropy_mean_weights_ii() -> None:	CODE
LOW	…x/backend/test/case/node/softmax_cross_entropy_loss.py	474	def export_softmaxcrossentropy_mean_weights_ii_log_prob() -> None:	CODE
LOW	…x/backend/test/case/node/softmax_cross_entropy_loss.py	509	def export_softmaxcrossentropy_mean_no_weights_ii() -> None:	CODE
LOW	…x/backend/test/case/node/softmax_cross_entropy_loss.py	538	def export_softmaxcrossentropy_mean_no_weights_ii_log_prob() -> None:	CODE
LOW	onnx/backend/test/case/node/max.py	51	def export_max_all_numeric_types() -> None:	CODE
LOW	onnx/backend/test/case/node/reducesum.py	89	def export_default_axes_keepdims() -> None:	CODE
LOW	onnx/backend/test/case/node/reducesum.py	124	def export_negative_axes_keepdims() -> None:	CODE
LOW	onnx/backend/test/case/node/reducesum.py	161	def export_empty_axes_input_noop() -> None:	CODE
LOW	onnx/backend/test/case/node/reducesum.py	225	def export_non_reduced_axis_zero() -> None:	CODE
LOW	onnx/backend/test/case/node/sequence_map.py	263	def export_sequence_map_extract_shapes(): # type: () -> None	CODE
LOW	onnx/backend/test/case/node/gather.py	72	def export_gather_negative_indices() -> None:	CODE
LOW	onnx/backend/test/case/node/attention.py	70	def export_attention_diff_head_sizes() -> None:	CODE
LOW	onnx/backend/test/case/node/attention.py	112	def export_attention_gqa_scaled() -> None:	CODE
LOW	onnx/backend/test/case/node/attention.py	136	def export_attention_diff_head_sizes_scaled() -> None:	CODE
LOW	onnx/backend/test/case/node/attention.py	183	def export_attention_gqa_causal() -> None:	CODE
LOW	onnx/backend/test/case/node/attention.py	206	def export_attention_diff_head_sizes_causal() -> None:	CODE
LOW	onnx/backend/test/case/node/attention.py	234	def export_attention_attn_mask() -> None:	CODE
LOW	onnx/backend/test/case/node/attention.py	378	def export_attention_attn_mask_bool() -> None:	CODE
948 more matches not shown…

Unused Imports868 hits · 734 pts

Severity	File	Line	Context
LOW	tools/gen_coverage_report.py	6	CODE
LOW	tools/pretty_clang_tidy.py	5	CODE
LOW	tools/check_namespace.py	10	CODE
LOW	onnx/model_container.py	8	CODE
LOW	onnx/gen_proto.py	6	CODE
LOW	onnx/version.py	10	CODE
LOW	onnx/version.py	14	CODE
LOW	onnx/compose.py	4	CODE
LOW	onnx/__init__.py	5	CODE
LOW	onnx/__init__.py	82	CODE
LOW	onnx/__init__.py	88	CODE
LOW	onnx/__init__.py	88	CODE
LOW	onnx/__init__.py	88	CODE
LOW	onnx/__init__.py	88	CODE
LOW	onnx/__init__.py	88	CODE
LOW	onnx/__init__.py	88	CODE
LOW	onnx/__init__.py	88	CODE
LOW	onnx/__init__.py	88	CODE
LOW	onnx/__init__.py	88	CODE
LOW	onnx/__init__.py	88	CODE
LOW	onnx/__init__.py	88	CODE
LOW	onnx/__init__.py	88	CODE
LOW	onnx/__init__.py	88	CODE
LOW	onnx/__init__.py	88	CODE
LOW	onnx/__init__.py	88	CODE
LOW	onnx/__init__.py	88	CODE
LOW	onnx/__init__.py	88	CODE
LOW	onnx/__init__.py	88	CODE
LOW	onnx/__init__.py	88	CODE
LOW	onnx/__init__.py	88	CODE
LOW	onnx/__init__.py	88	CODE
LOW	onnx/__init__.py	88	CODE
LOW	onnx/__init__.py	88	CODE
LOW	onnx/__init__.py	88	CODE
LOW	onnx/__init__.py	88	CODE
LOW	onnx/__init__.py	88	CODE
LOW	onnx/__init__.py	88	CODE
LOW	onnx/__init__.py	88	CODE
LOW	onnx/__init__.py	88	CODE
LOW	onnx/__init__.py	88	CODE
LOW	onnx/__init__.py	125	CODE
LOW	onnx/__init__.py	125	CODE
LOW	onnx/__init__.py	126	CODE
LOW	onnx/__init__.py	126	CODE
LOW	onnx/__init__.py	126	CODE
LOW	onnx/__init__.py	130	CODE
LOW	onnx/__init__.py	130	CODE
LOW	onnx/__init__.py	130	CODE
LOW	onnx/__init__.py	130	CODE
LOW	onnx/__init__.py	130	CODE
LOW	onnx/__init__.py	130	CODE
LOW	onnx/__init__.py	130	CODE
LOW	onnx/__init__.py	130	CODE
LOW	onnx/__init__.py	130	CODE
LOW	onnx/__init__.py	130	CODE
LOW	onnx/__init__.py	130	CODE
LOW	onnx/parser.py	4	CODE
LOW	onnx/shape_inference.py	10	CODE
LOW	onnx/utils.py	4	CODE
LOW	onnx/numpy_helper.py	4	CODE
808 more matches not shown…

Cross-File Repetition46 hits · 230 pts

Severity	File	Snippet	Context
HIGH	onnx/backend/test/case/node/flexattention.py	test flexattention with different head sizes for q/k vs v.	STRING
HIGH	docs/TestCoverage.md	test flexattention with different head sizes for q/k vs v.	STRING
HIGH	tests/python/shape_inference_test.py	test flexattention with different head sizes for q/k vs v.	STRING
HIGH	onnx/backend/test/case/node/averagepool.py	input_shape: [1, 1, 5, 5] output_shape: [1, 1, 5, 5] pad_shape: [4, 4] -> [2, 2, 2, 2] by axis	STRING
HIGH	onnx/backend/test/case/node/maxpool.py	input_shape: [1, 1, 5, 5] output_shape: [1, 1, 5, 5] pad_shape: [4, 4] -> [2, 2, 2, 2] by axis	STRING
HIGH	docs/TestCoverage.md	input_shape: [1, 1, 5, 5] output_shape: [1, 1, 5, 5] pad_shape: [4, 4] -> [2, 2, 2, 2] by axis	STRING
HIGH	onnx/backend/test/case/node/averagepool.py	input_shape: [1, 1, 5, 5] output_shape: [1, 1, 2, 2]	STRING
HIGH	onnx/backend/test/case/node/maxpool.py	input_shape: [1, 1, 5, 5] output_shape: [1, 1, 2, 2]	STRING
HIGH	docs/TestCoverage.md	input_shape: [1, 1, 5, 5] output_shape: [1, 1, 2, 2]	STRING
HIGH	onnx/backend/test/case/node/averagepool.py	input_shape: [1, 1, 5, 5] output_shape: [1, 1, 3, 3] pad_shape: [2, 2] -> [1, 1, 1, 1] by axis	STRING
HIGH	onnx/backend/test/case/node/maxpool.py	input_shape: [1, 1, 5, 5] output_shape: [1, 1, 3, 3] pad_shape: [2, 2] -> [1, 1, 1, 1] by axis	STRING
HIGH	docs/TestCoverage.md	input_shape: [1, 1, 5, 5] output_shape: [1, 1, 3, 3] pad_shape: [2, 2] -> [1, 1, 1, 1] by axis	STRING
HIGH	onnx/backend/test/case/node/averagepool.py	input_shape: [1, 3, 32, 32, 32] output_shape: [1, 3, 31, 31, 31]	STRING
HIGH	onnx/backend/test/case/node/maxpool.py	input_shape: [1, 3, 32, 32, 32] output_shape: [1, 3, 31, 31, 31]	STRING
HIGH	onnx/backend/test/case/node/lppool.py	input_shape: [1, 3, 32, 32, 32] output_shape: [1, 3, 31, 31, 31]	STRING
HIGH	docs/TestCoverage.md	input_shape: [1, 3, 32, 32, 32] output_shape: [1, 3, 31, 31, 31]	STRING
HIGH	onnx/backend/test/case/node/averagepool.py	input_shape: [1, 3, 32, 32] output_shape: [1, 3, 32, 32] pad_shape: [1, 1] -> [1, 0, 1, 0] by axis	STRING
HIGH	onnx/backend/test/case/node/maxpool.py	input_shape: [1, 3, 32, 32] output_shape: [1, 3, 32, 32] pad_shape: [1, 1] -> [1, 0, 1, 0] by axis	STRING
HIGH	onnx/backend/test/case/node/lppool.py	input_shape: [1, 3, 32, 32] output_shape: [1, 3, 32, 32] pad_shape: [1, 1] -> [1, 0, 1, 0] by axis	STRING
HIGH	docs/TestCoverage.md	input_shape: [1, 3, 32, 32] output_shape: [1, 3, 32, 32] pad_shape: [1, 1] -> [1, 0, 1, 0] by axis	STRING
HIGH	onnx/backend/test/case/node/averagepool.py	input_shape: [1, 3, 28, 28] output_shape: [1, 3, 30, 30] pad_shape: [4, 4] -> [2, 2, 2, 2] by axis	STRING
HIGH	onnx/backend/test/case/node/maxpool.py	input_shape: [1, 3, 28, 28] output_shape: [1, 3, 30, 30] pad_shape: [4, 4] -> [2, 2, 2, 2] by axis	STRING
HIGH	onnx/backend/test/case/node/lppool.py	input_shape: [1, 3, 28, 28] output_shape: [1, 3, 30, 30] pad_shape: [4, 4] -> [2, 2, 2, 2] by axis	STRING
HIGH	docs/TestCoverage.md	input_shape: [1, 3, 28, 28] output_shape: [1, 3, 30, 30] pad_shape: [4, 4] -> [2, 2, 2, 2] by axis	STRING
HIGH	onnx/backend/test/case/node/averagepool.py	input_shape: [1, 3, 32, 32] output_shape: [1, 3, 10, 10]	STRING
HIGH	onnx/backend/test/case/node/maxpool.py	input_shape: [1, 3, 32, 32] output_shape: [1, 3, 10, 10]	STRING
HIGH	onnx/backend/test/case/node/lppool.py	input_shape: [1, 3, 32, 32] output_shape: [1, 3, 10, 10]	STRING
HIGH	docs/TestCoverage.md	input_shape: [1, 3, 32, 32] output_shape: [1, 3, 10, 10]	STRING
HIGH	onnx/backend/test/case/node/averagepool.py	input_shape: [1, 1, 4, 4, 4] output_shape: [1, 1, 2, 2, 2]	STRING
HIGH	onnx/backend/test/case/node/maxpool.py	input_shape: [1, 1, 4, 4, 4] output_shape: [1, 1, 2, 2, 2]	STRING
HIGH	onnx/backend/test/case/node/lppool.py	input_shape: [1, 1, 4, 4, 4] output_shape: [1, 1, 2, 2, 2]	STRING
HIGH	docs/TestCoverage.md	input_shape: [1, 1, 4, 4, 4] output_shape: [1, 1, 2, 2, 2]	STRING
HIGH	onnx/backend/test/case/node/averagepool.py	input_shape: [1, 1, 2, 2] output_shape: [1, 1, 1, 1]	STRING
HIGH	docs/TestCoverage.md	input_shape: [1, 1, 2, 2] output_shape: [1, 1, 1, 1]	STRING
HIGH	onnx/backend/test/case/node/maxpool.py	input_shape: [1, 1, 2, 2] output_shape: [1, 1, 1, 1]	STRING
HIGH	onnx/backend/test/case/node/gru.py	run a forward pass of the rnn. assumes that the num_directions axis has been squeezed out of the inputs. (and returns y,	STRING
HIGH	onnx/reference/ops/op_gru.py	run a forward pass of the rnn. assumes that the num_directions axis has been squeezed out of the inputs. (and returns y,	STRING
HIGH	onnx/backend/test/case/node/rnn.py	run a forward pass of the rnn. assumes that the num_directions axis has been squeezed out of the inputs. (and returns y,	STRING
HIGH	onnx/reference/ops/op_rnn.py	run a forward pass of the rnn. assumes that the num_directions axis has been squeezed out of the inputs. (and returns y,	STRING
HIGH	onnx/reference/ops/experimental/_op_list.py	loads the implemented for a specified operator. args: domain: domain op_type: operator type version: requested version c	STRING
HIGH	onnx/reference/ops/aionnx_preview_training/_op_list.py	loads the implemented for a specified operator. args: domain: domain op_type: operator type version: requested version c	STRING
HIGH	onnx/reference/ops/aionnx_preview/_op_list.py	loads the implemented for a specified operator. args: domain: domain op_type: operator type version: requested version c	STRING
HIGH	onnx/reference/ops/aionnxml/_op_list.py	loads the implemented for a specified operator. args: domain: domain op_type: operator type version: requested version c	STRING
HIGH	tests/python/checker_test.py	<ir_version: 8, opset_import: [ "" : 17, "local" : 1 ]> agraph (float[n] x) => (float[n] y) { y = local.foo (x) } <opset	STRING
HIGH	tests/python/inliner_test.py	<ir_version: 8, opset_import: [ "" : 17, "local" : 1 ]> agraph (float[n] x) => (float[n] y) { y = local.foo (x) } <opset	STRING
HIGH	tests/python/shape_inference_test.py	<ir_version: 8, opset_import: [ "" : 17, "local" : 1 ]> agraph (float[n] x) => (float[n] y) { y = local.foo (x) } <opset	STRING

Deep Nesting99 hits · 98 pts

Severity	File	Line	Context
LOW	tools/check_namespace.py	15	CODE
LOW	onnx/gen_proto.py	41	CODE
LOW	onnx/compose.py	91	CODE
LOW	onnx/compose.py	446	CODE
LOW	onnx/compose.py	569	CODE
LOW	onnx/compose.py	221	CODE
LOW	onnx/numpy_helper.py	321	CODE
LOW	onnx/numpy_helper.py	400	CODE
LOW	onnx/numpy_helper.py	562	CODE
LOW	onnx/helper.py	86	CODE
LOW	onnx/helper.py	365	CODE
LOW	onnx/helper.py	508	CODE
LOW	onnx/helper.py	570	CODE
LOW	onnx/helper.py	608	CODE
LOW	onnx/helper.py	787	CODE
LOW	onnx/helper.py	850	CODE
LOW	onnx/helper.py	993	CODE
LOW	onnx/helper.py	1231	CODE
LOW	onnx/external_data_helper.py	207	CODE
LOW	onnx/external_data_helper.py	61	CODE
LOW	onnx/tools/replace_constants.py	199	CODE
LOW	onnx/backend/test/stat_coverage.py	40	CODE
LOW	onnx/backend/test/stat_coverage.py	152	CODE
LOW	onnx/backend/test/cmd_tools.py	18	CODE
LOW	onnx/backend/test/runner/__init__.py	123	CODE
LOW	onnx/backend/test/runner/__init__.py	254	CODE
LOW	onnx/backend/test/runner/__init__.py	341	CODE
LOW	onnx/backend/test/runner/__init__.py	517	CODE
LOW	onnx/backend/test/runner/__init__.py	367	CODE
LOW	onnx/backend/test/loader/__init__.py	25	CODE
LOW	onnx/backend/test/case/node/cast.py	27	CODE
LOW	onnx/backend/test/case/node/cast.py	321	CODE
LOW	onnx/backend/test/case/node/__init__.py	34	CODE
LOW	onnx/backend/test/case/node/__init__.py	261	CODE
LOW	onnx/backend/test/case/node/scatternd.py	13	CODE
LOW	onnx/backend/test/case/node/scatterelements.py	19	CODE
LOW	onnx/backend/test/case/node/castlike.py	23	CODE
LOW	onnx/backend/test/report/coverage.py	147	CODE
LOW	onnx/reference/op_run.py	157	CODE
LOW	onnx/reference/op_run.py	240	CODE
LOW	onnx/reference/op_run.py	395	CODE
LOW	onnx/reference/shape_annotation_checker.py	59	CODE
LOW	onnx/reference/reference_evaluator.py	209	CODE
LOW	onnx/reference/reference_evaluator.py	411	CODE
LOW	onnx/reference/reference_evaluator.py	579	CODE
LOW	onnx/reference/ops_optimized/op_conv_optimized.py	50	CODE
LOW	onnx/reference/ops/op_scatter_elements.py	11	CODE
LOW	onnx/reference/ops/op_constant.py	71	CODE
LOW	onnx/reference/ops/op_roi_align.py	28	CODE
LOW	onnx/reference/ops/op_roi_align.py	117	CODE
LOW	onnx/reference/ops/op_grid_sample.py	286	CODE
LOW	onnx/reference/ops/op_sequence_insert.py	41	CODE
LOW	onnx/reference/ops/op_conv.py	11	CODE
LOW	onnx/reference/ops/_op_common_pool.py	64	CODE
LOW	onnx/reference/ops/_op_common_pool.py	119	CODE
LOW	onnx/reference/ops/op_affine_grid.py	11	CODE
LOW	onnx/reference/ops/op_non_max_suppression.py	176	CODE
LOW	onnx/reference/ops/op_pool_common.py	163	CODE
LOW	onnx/reference/ops/op_deform_conv.py	12	CODE
LOW	onnx/reference/ops/op_prelu.py	12	CODE
39 more matches not shown…

Over-Commented Block92 hits · 90 pts

Severity	File	Line	Snippet	Context
LOW	.lintrunner.toml	1	# Configuration for lintrunner https://github.com/suo/lintrunner	COMMENT
LOW	.lintrunner.toml	21	#	COMMENT
LOW	onnx/onnx_pb.h	1	// Copyright (c) ONNX Project Contributors	COMMENT
LOW	onnx/onnx_pb.h	21	#define ONNX_EXPORT __attribute__((__visibility__("default")))	COMMENT
LOW	onnx/onnx_pb.h	41	//	COMMENT
LOW	onnx/checker.h	1	// Copyright (c) ONNX Project Contributors	COMMENT
LOW	onnx/onnx-operators_pb.h	1	// Copyright (c) ONNX Project Contributors	COMMENT
LOW	onnx/proto_utils.h	1	// Copyright (c) ONNX Project Contributors	COMMENT
LOW	onnx/version_converter/convert.h	1	// Copyright (c) ONNX Project Contributors	COMMENT
LOW	onnx/version_converter/convert.h	21	#include "onnx/version_converter/adapters/Attention_25_24.h"	COMMENT
LOW	onnx/version_converter/convert.h	41	#include "onnx/version_converter/adapters/q_dq_21_20.h"	COMMENT
LOW	onnx/version_converter/BaseConverter.h	1	// Copyright (c) ONNX Project Contributors	COMMENT
LOW	onnx/version_converter/adapters/q_dq_21_20.h	1	// Copyright (c) ONNX Project Contributors	COMMENT
LOW	onnx/version_converter/adapters/type_restriction.h	1	// Copyright (c) ONNX Project Contributors	COMMENT
LOW	onnx/version_converter/adapters/adapter.h	1	// Copyright (c) ONNX Project Contributors	COMMENT
LOW	…_converter/adapters/broadcast_backward_compatibility.h	1	// Copyright (c) ONNX Project Contributors	COMMENT
LOW	onnx/version_converter/adapters/range_27_26.h	1	// Copyright (c) ONNX Project Contributors	COMMENT
LOW	onnx/version_converter/adapters/upsample_9_8.h	1	// Copyright (c) ONNX Project Contributors	COMMENT
LOW	onnx/version_converter/adapters/scatter_16_15.h	1	// Copyright (c) ONNX Project Contributors	COMMENT
LOW	…nx/version_converter/adapters/extend_supported_types.h	1	// Copyright (c) ONNX Project Contributors	COMMENT
LOW	onnx/version_converter/adapters/transformers.h	1	// Copyright (c) ONNX Project Contributors	COMMENT
LOW	onnx/tools/__init__.py	1	# Copyright (c) ONNX Project Contributors	COMMENT
LOW	onnx/frontend/__init__.py	1	# Copyright (c) ONNX Project Contributors	COMMENT
LOW	onnx/bin/__init__.py	1	# Copyright (c) ONNX Project Contributors	COMMENT
LOW	onnx/shape_inference/implementation.h	1	// Copyright (c) ONNX Project Contributors	COMMENT
LOW	onnx/shape_inference/implementation.h	541	const std::unordered_map<std::string, int>& func_opset_imports,	COMMENT
LOW	onnx/backend/__init__.py	1	# Copyright (c) ONNX Project Contributors	COMMENT
LOW	onnx/backend/test/case/node/unique.py	161	# print(y)	COMMENT
LOW	onnx/backend/test/case/node/__init__.py	241	in {"ai.onnx.training version", "ai.onnx.preview.training"}	COMMENT
LOW	onnx/backend/test/case/node/resize.py	401	)	COMMENT
LOW	onnx/backend/test/case/node/resize.py	541		COMMENT
LOW	onnx/backend/test/case/node/resize.py	641	# 3.96160918 4.50364964 4.75976814]	COMMENT
LOW	onnx/backend/test/case/node/resize.py	741	# [ 2.625 3.03125 3.625 4.125 4.625 5.21875 5.625	COMMENT
LOW	onnx/backend/test/case/node/trilu.py	141	outputs=["y"],	COMMENT
LOW	onnx/backend/test/case/node/trilu.py	201	x = np.random.randint(10, size=(3, 1, 5)).astype(np.int64)	COMMENT
LOW	onnx/backend/test/case/node/trilu.py	361		COMMENT
LOW	onnx/backend/test/case/node/trilu.py	421	x = np.random.randint(10, size=(3, 1, 5)).astype(np.int64)	COMMENT
LOW	onnx/common/file_utils.h	1	// Copyright (c) ONNX Project Contributors	COMMENT
LOW	onnx/common/path.h	1	// Copyright (c) ONNX Project Contributors	COMMENT
LOW	onnx/common/ir.h	1	// Copyright (c) ONNX Project Contributors	COMMENT
LOW	onnx/common/ir.h	21	#include <unordered_map>	COMMENT
LOW	onnx/common/ir.h	641	// with an input which is topologically before the current node.	COMMENT
LOW	onnx/common/ir.h	661	// Replace the input of 'this' at position 'i' with	COMMENT
LOW	onnx/common/ir.h	701	// and unattached: %5 = h(%1)	COMMENT
LOW	onnx/common/ir.h	741	removeFromList();	COMMENT
LOW	onnx/common/ir.h	761	//	COMMENT
LOW	onnx/common/ir.h	961	// having corner cases where the list is empty.	COMMENT
LOW	onnx/common/ir.h	1001	// (called from Attributes<Node> after set/removeAttribute/copyAttributes)	COMMENT
LOW	onnx/common/scoped_resource.h	1	// Copyright (c) ONNX Project Contributors	COMMENT
LOW	onnx/common/scoped_resource.h	21		COMMENT
LOW	onnx/common/graph_node_list.h	1	// Copyright (c) ONNX Project Contributors	COMMENT
LOW	onnx/common/graph_node_list.h	21	// things, you will have to do some refactoring.	COMMENT
LOW	onnx/common/common.h	21		COMMENT
LOW	onnx/common/safe_math.h	1	// Copyright (c) ONNX Project Contributors	COMMENT
LOW	onnx/common/proto_util.h	1	// Copyright (c) ONNX Project Contributors	COMMENT
LOW	onnx/common/tensor.h	1	// Copyright (c) ONNX Project Contributors	COMMENT
LOW	onnx/common/array_ref.h	1	// Copyright (c) ONNX Project Contributors	COMMENT
LOW	onnx/common/array_ref.h	21	// removed a bunch of slice variants for simplicity...	COMMENT
LOW	onnx/common/model_helpers.h	1	// Copyright (c) ONNX Project Contributors	COMMENT
LOW	onnx/inliner/inliner.h	1	// Copyright (c) ONNX Project Contributors	COMMENT
32 more matches not shown…

Decorative Section Separators12 hits · 36 pts

Severity	File	Line	Snippet	Context
MEDIUM	onnx/backend/test/case/node/gridsample.py	303	# ============================================================================	COMMENT
MEDIUM	onnx/backend/test/case/node/linear_attention.py	59	# ------------------------------------------------------------------	COMMENT
MEDIUM	onnx/backend/test/case/node/linear_attention.py	61	# ------------------------------------------------------------------	COMMENT
MEDIUM	onnx/backend/test/case/node/linear_attention.py	260	# ------------------------------------------------------------------	COMMENT
MEDIUM	onnx/backend/test/case/node/linear_attention.py	262	# ------------------------------------------------------------------	COMMENT
MEDIUM	onnx/backend/test/case/node/linear_attention.py	329	# ------------------------------------------------------------------	COMMENT
MEDIUM	onnx/backend/test/case/node/linear_attention.py	331	# ------------------------------------------------------------------	COMMENT
MEDIUM	onnx/backend/test/case/node/linear_attention.py	437	# ------------------------------------------------------------------	COMMENT
MEDIUM	onnx/backend/test/case/node/linear_attention.py	439	# ------------------------------------------------------------------	COMMENT
MEDIUM	onnx/backend/test/case/node/linear_attention.py	509	# ------------------------------------------------------------------	COMMENT
MEDIUM	onnx/backend/test/case/node/linear_attention.py	511	# ------------------------------------------------------------------	COMMENT
MEDIUM	docs/TestCoverage.md	12405	# ============================================================================	COMMENT

Self-Referential Comments16 hits · 34 pts

Severity	File	Line	Snippet	Context
MEDIUM	onnx/gen_proto.py	214	# This file is generated by 'gen_proto.py'. DO NOT EDIT!	COMMENT
MEDIUM	onnx/numpy_helper.py	130	# Create a 1D copy	COMMENT
MEDIUM	onnx/numpy_helper.py	174	# Create a 1D copy	COMMENT
MEDIUM	onnx/backend/test/case/node/image_decoder.py	16	# Create an empty RGB image	COMMENT
MEDIUM	onnx/backend/test/case/node/_image_decoder_data.py	5	# This file contains freeze NumPy array for ImageDecoder backend test's input/output	COMMENT
MEDIUM	onnx/reference/ops/op_pool_common.py	176	"""This function is used to calculate the pooling result of a padded tensor	STRING
MEDIUM	tests/python/version_converter_test.py	2580	# Create a model with external data	COMMENT
MEDIUM	tests/python/external_data_test.py	950	# Create a target file and replace external data with a symlink to it	COMMENT
MEDIUM	tests/python/external_data_test.py	1000	# Create a model directory with a real subdir for saving	COMMENT
MEDIUM	tests/python/external_data_test.py	1046	# Create a hardlink to the external data file	COMMENT
MEDIUM	tests/python/shape_inference_test.py	6406	# Create a simple If node where the 'then' subgraph adds to the current value, and the 'else' subgraph	STRING
MEDIUM	tests/python/shape_inference_test.py	6428	# Create a simple If node where the 'then' subgraph adds to the current value, and the 'else' subgraph	STRING
MEDIUM	tests/python/shape_inference_test.py	6448	# Create a simple If node where the 'then' subgraph adds to the current value, and the 'else' subgraph	STRING
MEDIUM	tests/python/shape_inference_test.py	8326	# Create a TensorProto with incorrect data type for `delta`.	STRING
MEDIUM	tests/python/shape_inference_test.py	8352	# Create a TensorProto with empty raw data for `delta`.	STRING
MEDIUM	tests/python/training_tool_test.py	69	# Create a model with both inference and training information.	COMMENT

AI Slop Vocabulary10 hits · 30 pts

Severity	File	Line	Snippet	Context
MEDIUM	pyproject.toml	286	"PERF203", # try-except in loop is unavoidable in the version-converter harness	CODE
MEDIUM	onnx/fuzz/make_seed_corpus.py	187	# ModelProto with no trailing bytes: the harness's raw path passes the full	STRING
MEDIUM	onnx/fuzz/make_seed_corpus.py	334	# harness's raw path: a 4-byte big-endian length prefix for m1's serialized	COMMENT
MEDIUM	onnx/fuzz/make_seed_corpus.py	337	# The harness derives the io_map from m1's output names and m2's input names,	COMMENT
MEDIUM	onnx/fuzz/make_seed_corpus.py	518	# serialized ModelProto with no trailing bytes, so the harness's raw path	COMMENT
MEDIUM	onnx/reference/ops/op_lp_pool.py	24	# utilize AvgPool the same fashion Pytorch does. Note that there is a difference in computation.	COMMENT
MEDIUM⚡	.github/workflows/fuzz.yml	8	# (see google/oss-fuzz#15382) — it exists to catch harness regressions	COMMENT
MEDIUM⚡	.github/workflows/fuzz.yml	9	# (import errors, API drift, a harness file going missing) at PR time	COMMENT
MEDIUM⚡	.github/workflows/fuzz.yml	14	# for the harness/toggle-byte details.	COMMENT
MEDIUM	.github/workflows/fuzz.yml	32	- cron: '0 6 * * *' # nightly: run each harness for longer than the PR smoke test	CODE

AI Structural Patterns26 hits · 26 pts

Severity	File	Line	Context
LOW	onnx/compose.py	300	CODE
LOW	onnx/compose.py	446	CODE
LOW	onnx/compose.py	569	CODE
LOW	onnx/reference/ops/op_attention.py	120	CODE
LOW	onnx/reference/ops/op_attention.py	394	CODE
LOW	onnx/reference/ops/op_roi_align.py	12	CODE
LOW	onnx/reference/ops/op_linear_attention.py	23	CODE
LOW	onnx/reference/ops/op_scan.py	114	CODE
LOW	onnx/reference/ops/op_gru.py	136	CODE
LOW	onnx/reference/ops/op_lstm.py	147	CODE
LOW	onnx/reference/ops/op_lp_pool.py	12	CODE
LOW	onnx/reference/ops/op_conv_integer.py	13	CODE
LOW	onnx/reference/ops/op_rnn.py	104	CODE
LOW	onnx/reference/ops/op_deform_conv.py	149	CODE
LOW	onnx/reference/ops/op_attribute_has_value.py	12	CODE
LOW	onnx/reference/ops/op_resize.py	580	CODE
LOW	onnx/reference/ops/op_tfidf_vectorizer.py	237	CODE
LOW	onnx/reference/ops/op_tfidf_vectorizer.py	300	CODE
LOW	onnx/reference/ops/op_conv_transpose.py	13	CODE
LOW	…/reference/ops/aionnxml/op_tree_ensemble_classifier.py	19	CODE
LOW	onnx/reference/ops/aionnxml/op_svm_classifier.py	223	CODE
LOW	…x/reference/ops/aionnxml/op_tree_ensemble_regressor.py	15	CODE
LOW	onnx/reference/ops/aionnxml/op_svm_regressor.py	13	CODE
LOW	onnx/reference/ops/aionnxml/op_label_encoder.py	12	CODE
LOW	tests/python/compose_test.py	740	CODE
LOW	…on/version_converter/automatic_conversion_test_base.py	44	CODE

Docstring Block Structure4 hits · 20 pts

Severity	File	Line	Snippet	Context
HIGH	onnx/serialization.py	64	Get a serializer for a format. Args: fmt: The format to get a serializer for. Returns:	STRING
HIGH	onnx/reference/op_run.py	396	Calls method ``_run``, catches exceptions, displays a longer error message. Args: *args: in	STRING
HIGH	onnx/reference/reference_evaluator.py	588	Executes the onnx model. Args: output_names: requested outputs by names, None for all f	STRING
HIGH	onnx/reference/ops/op_resize.py	135	Return the n nearest indexes to x among `[0, limit)`, prefer the indexes smaller than x. As a result, the ratio	STRING

Excessive Try-Catch Wrapping17 hits · 19 pts

Severity	File	Line	Snippet	Context
LOW	onnx/backend/test/runner/__init__.py	49	except Exception: # noqa: PERF203	CODE
LOW	onnx/backend/test/runner/__init__.py	249	except Exception as e:	CODE
LOW	onnx/fuzz/fuzz_version_converter.py	40	except Exception:	CODE
LOW	onnx/fuzz/fuzz_version_converter.py	46	except Exception:	CODE
MEDIUM	onnx/fuzz/fuzz_version_converter.py	37	def TestOneInput(data):	CODE
LOW	onnx/fuzz/fuzz_compose.py	133	except Exception:	CODE
LOW	onnx/fuzz/fuzz_parser.py	17	except Exception:	CODE
MEDIUM	onnx/fuzz/fuzz_parser.py	13	def TestOneInput(data):	CODE
LOW	onnx/fuzz/fuzz_model_loader.py	20	except Exception:	CODE
MEDIUM	onnx/fuzz/fuzz_model_loader.py	13	def TestOneInput(data):	CODE
LOW	onnx/fuzz/fuzz_shape_inference.py	175	except Exception:	CODE
LOW	onnx/fuzz/fuzz_checker.py	16	except Exception:	CODE
MEDIUM	onnx/fuzz/fuzz_checker.py	13	def TestOneInput(data):	CODE
LOW⚡	tests/python/reference_evaluator_test.py	260	except Exception as e:	STRING
LOW	docs/docsgen/source/intro/python.md	1168	except Exception as e:	STRING
LOW	docs/docsgen/source/intro/python.md	1585	except Exception as e:	STRING
LOW	docs/docsgen/source/intro/python.md	1592	except Exception as e:	STRING

Modern Structural Boilerplate19 hits · 19 pts

Severity	File	Line	Snippet	Context
LOW	onnx/version.py	24	__all__ = ["version", "git_version"]	CODE
LOW	onnx/__init__.py	7	__all__ = [	CODE
LOW	onnx/helper.py	361	def set_model_props(model: ModelProto, dict_value: dict[str, str]) -> None:	CODE
LOW	onnx/serialization.py	9	__all__ = [	CODE
LOW	onnx/checker.py	8	__all__ = [	CODE
LOW	onnx/tools/update_model_dims.py	62	def update_dim(tensor: ValueInfoProto, dim: Any, j: int, name: str) -> None:	CODE
LOW	onnx/backend/test/__init__.py	6	__all__ = ["BackendTest"]	CODE
LOW	onnx/reference/__init__.py	6	__all__ = ["ReferenceEvaluator"]	CODE
LOW	onnx/reference/ops_optimized/__init__.py	10	__all__ = ["Conv", "optimized_operators"]	CODE
LOW	onnx/reference/ops/__init__.py	6	__all__ = ["load_op"]	CODE
LOW	onnx/reference/ops/_op_list.py	15	__all__ = [	CODE
LOW	onnx/reference/ops/experimental/__init__.py	6	__all__ = ["load_op"]	CODE
LOW	onnx/reference/ops/aionnx_preview_training/__init__.py	6	__all__ = ["load_op", "OpRunTraining"]	CODE
LOW	onnx/reference/ops/aionnx_preview_training/_op_list.py	6	__all__ = [	CODE
LOW	onnx/reference/ops/aionnx_preview/__init__.py	6	__all__ = ["load_op"]	CODE
LOW	onnx/reference/ops/aionnx_preview/_op_list.py	6	__all__ = [	CODE
LOW	onnx/reference/ops/aionnxml/__init__.py	4	__all__ = ["load_op"]	CODE
LOW	onnx/reference/ops/aionnxml/_op_list.py	8	__all__ = [	CODE
LOW	onnx/defs/__init__.py	6	__all__ = [	CODE

Structural Annotation Overuse8 hits · 15 pts

Severity	File	Line	Snippet	Context
LOW⚡	onnx/reference/ops/op_causal_conv_with_state.py	41	# Step 1: build the left-padded input (B, C, L + k - 1).	COMMENT
LOW⚡	onnx/reference/ops/op_causal_conv_with_state.py	48	# Step 2: depthwise Conv1d (group = channels, valid padding).	COMMENT
LOW⚡	onnx/reference/ops/op_causal_conv_with_state.py	61	# Step 3: optional fused SiLU/Swish activation.	COMMENT
LOW⚡	onnx/reference/ops/op_causal_conv_with_state.py	68	# Step 4: present_state = last (k - 1) positions of the padded input.	COMMENT
LOW	docs/AddNewOp.md	43	### Step 1: Proposing a new operator/function	COMMENT
LOW	docs/AddNewOp.md	61	### Step 2: Submit PR	COMMENT
LOW	docs/AddNewOp.md	113	### Step 3: PR Review by Operators SIG	COMMENT
LOW	docs/AddNewOp.md	121	### Step 4: ONNX release	COMMENT

Verbosity Indicators4 hits · 9 pts

Severity	File	Line	Snippet	Context
LOW⚡	onnx/reference/ops/op_causal_conv_with_state.py	41	# Step 1: build the left-padded input (B, C, L + k - 1).	COMMENT
LOW⚡	onnx/reference/ops/op_causal_conv_with_state.py	48	# Step 2: depthwise Conv1d (group = channels, valid padding).	COMMENT
LOW⚡	onnx/reference/ops/op_causal_conv_with_state.py	61	# Step 3: optional fused SiLU/Swish activation.	COMMENT
LOW⚡	onnx/reference/ops/op_causal_conv_with_state.py	68	# Step 4: present_state = last (k - 1) positions of the padded input.	COMMENT

Redundant / Tautological Comments5 hits · 8 pts

Severity	File	Line	Snippet	Context
LOW	onnx/reference/ops/op_string_split.py	30	# Add padding to lists that are shorter than the maximum length	COMMENT
LOW	tests/python/training_tool_test.py	71	# Check if the inference-only part is correct.	COMMENT
LOW⚡	tests/python/basic_test.py	208	# Check if the version is correct.	COMMENT
LOW⚡	tests/python/basic_test.py	212	# Check if the __repr__ methods work without error	COMMENT
LOW	tests/python/compose_test.py	482	# Check if all inputs are prefixed even if input is not connected to node. e.g., 'unused_input'.	STRING

Cross-Language Confusion1 hit · 8 pts

Severity	File	Line	Snippet	Context
HIGH	onnx/reference/op_run.py	277	# default value is undefined, it depends on the inputs	COMMENT

AI Response Leakage1 hit · 8 pts

Severity	File	Line	Snippet	Context
HIGH	onnx/reference/ops/op_grid_sample.py	311	# In this implementation, we took the latter approach.	COMMENT

Modern AI Meta-Vocabulary1 hit · 2 pts

Severity	File	Line	Snippet	Context
MEDIUM	pyproject.toml	284	"PLR2004", # Magic numbers are fine in low-level fuzz scaffolding	CODE

Fake / Example Data1 hit · 1 pts

Severity	File	Line	Snippet	Context
LOW	docs/docsgen/source/_static/diff2html-ui-slim.min.js	1	!function(e,n){if("object"==typeof exports&&"object"==typeof module)module.exports=n();else if("function"==typeof define	CODE

Overly Generic Function Names1 hit · 0 pts

Severity	File	Line	Snippet	Context
LOW	tests/python/external_data_test.py	828	def test_function(self) -> None:	CODE

Analysis Overview

What These Metrics Mean

Score History

Severity Breakdown

Directory Score Breakdown

Pattern Findings