HIPS/autograd

15.4

Adjusted Score

15.4

Raw Score

100%

Time Factor

2026-07-13

Last Push

7.5K

Stars

Python

Language

15.4K

Lines of Code

120

Files

205

Pattern Hits

2026-07-14

Scan Date

0.03

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 3MEDIUM 4LOW 198

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 205 distinct pattern matches across 11 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 Identifiers108 hits · 118 pts

Severity	File	Line	Snippet	Context
LOW	autograd/builtins.py	121	def grad_sequence_extend_right(argnum, ans, args, kwargs):	CODE
LOW	autograd/builtins.py	134	def grad_sequence_extend_left(argnum, ans, args, kwargs):	CODE
LOW	autograd/core.py	379	def deprecated_defvjp_is_zero(primitive_fun):	CODE
LOW	autograd/core.py	408	def primitive_with_deprecation_warnings(f_raw):	CODE
LOW	autograd/numpy/numpy_wrapper.py	101	def _array_from_scalar_or_array(array_args, array_kwargs, scalar):	CODE
LOW	autograd/numpy/numpy_vjps.py	922	def array_from_args_gradmaker(argnum, ans, args, kwargs):	CODE
LOW	autograd/numpy/numpy_vjps.py	929	def array_from_scalar_or_array_gradmaker(ans, array_args, array_kwargs, scarray):	CODE
LOW	autograd/numpy/numpy_jvps.py	243	def fwd_grad_concatenate_args(argnum, g, ans, axis_args, kwargs):	CODE
LOW	autograd/scipy/stats/multivariate_normal.py	19	def generalized_outer_product(x):	CODE
LOW	tests/test_linalg.py	17	def check_symmetric_matrix_grads(fun, **grad_check_kwargs):	CODE
LOW	tests/test_linalg.py	204	def test_frobenius_norm_complex():	CODE
LOW	tests/test_linalg.py	222	def test_frobenius_norm_axis_complex():	CODE
LOW	tests/test_linalg.py	243	def test_vector_norm_ord_complex(size, ord):	CODE
LOW	tests/test_linalg.py	281	def test_norm_nuclear_complex():	CODE
LOW	tests/test_linalg.py	300	def test_norm_nuclear_axis_complex():	CODE
LOW	tests/test_linalg.py	332	def test_eigvalh_lower_broadcasting():	CODE
LOW	tests/test_linalg.py	343	def test_eigvalh_upper_broadcasting():	CODE
LOW	tests/test_linalg.py	357	def test_eigvalh_lower_complex():	CODE
LOW	tests/test_linalg.py	367	def test_eigvalh_upper_complex():	CODE
LOW	tests/test_linalg.py	420	def test_cholesky_reparameterization_trick():	CODE
LOW	tests/test_misc.py	96	def test_flatten_nodes_in_containers():	CODE
LOW⚡	tests/test_scipy.py	104	def test_norm_logpdf_broadcast():	CODE
LOW⚡	tests/test_scipy.py	107	def test_norm_logcdf_broadcast():	CODE
LOW⚡	tests/test_scipy.py	110	def test_norm_logsf_broadcast():	CODE
LOW⚡	tests/test_scipy.py	122	def test_poisson_cdf_broadcast():	CODE
LOW⚡	tests/test_scipy.py	125	def test_poisson_logpmf_broadcast():	CODE
LOW⚡	tests/test_scipy.py	128	def test_poisson_pmf_broadcast():	CODE
LOW	tests/test_scipy.py	200	def test_mvn_logpdf_broadcast():	CODE
LOW	tests/test_scipy.py	213	def normalized_dirichlet_logpdf(x, alpha):	CODE
LOW	tests/test_scipy.py	225	def test_dirichlet_logpdf_alpha():	CODE
LOW	tests/test_scipy.py	269	def test_convolve_generalization():	CODE
LOW⚡	tests/test_vspaces.py	49	def compatibility_of_scalar_mul_with_field_mul(x, a, b):	CODE
LOW⚡	tests/test_vspaces.py	52	def identity_element_of_scalar_mul(x):	CODE
LOW⚡	tests/test_vspaces.py	55	def distributivity_of_scalar_mul_wrt_vector_add(x, y, a):	CODE
LOW⚡	tests/test_vspaces.py	58	def distributivity_of_scalar_mul_wrt_scalar_add(x, a, b):	CODE
LOW⚡	tests/test_vspaces.py	65	def scalar_mul_preserves_vspace(x, a):	CODE
LOW	tests/test_vspaces.py	144	def test_array_vspace_complex():	CODE
LOW	tests/test_graphs.py	83	def test_mutating_outgrad_from_indexing():	CODE
LOW	tests/test_graphs.py	95	def test_complex_mutating_outgrad_from_indexing():	CODE
LOW	tests/test_graphs.py	109	def test_complex_separate_real_and_imaginary():	CODE
LOW	tests/test_graphs.py	132	def test_third_derivative_other_args():	CODE
LOW	tests/test_graphs.py	154	def test_singleton_array_output():	CODE
LOW	tests/test_graphs.py	184	def test_assignment_raises_error():	CODE
LOW	tests/numpy_utils.py	54	def binary_ufunc_check_no_same_args(fun, lims_A=[-2, 2], lims_B=[-2, 2], test_complex=True, **kwargs):	CODE
LOW⚡	tests/test_jacobian.py	9	def test_jacobian_against_grad():	CODE
LOW⚡	tests/test_jacobian.py	15	def test_jacobian_scalar_to_vector():	CODE
LOW⚡	tests/test_jacobian.py	21	def test_jacobian_against_stacked_grads():	CODE
LOW	tests/test_jacobian.py	37	def test_jacobian_higher_order():	CODE
LOW	tests/test_fft.py	270	def test_rfftn_odd_not_implemented():	CODE
LOW	tests/test_ufunc_dispatch.py	22	def test_grad_through_dataarray_binary_ufunc():	CODE
LOW	tests/test_ufunc_dispatch.py	35	def test_grad_through_dataarray_reversed_operand():	CODE
LOW	tests/test_ufunc_dispatch.py	48	def test_grad_through_dataarray_unary_ufunc():	CODE
LOW	tests/test_ufunc_dispatch.py	62	def test_grad_through_dataarray_transcendental_ufunc():	CODE
LOW	tests/test_ufunc_dispatch.py	76	def test_grad_through_two_dataarrays_holding_boxes():	CODE
LOW	tests/test_ufunc_dispatch.py	92	def test_grad_through_dataarray_box_plus_plain_dataarray():	CODE
LOW	tests/test_ufunc_dispatch.py	106	def test_grad_through_dataarray_box_times_plain_ndarray():	CODE
LOW	tests/test_ufunc_dispatch.py	120	def test_grad_through_dataarray_box_times_python_scalar():	CODE
LOW	tests/test_ufunc_dispatch.py	134	def test_grad_through_dataarray_division():	CODE
LOW	tests/test_ufunc_dispatch.py	147	def test_grad_through_dataarray_exp():	CODE
LOW	tests/test_ufunc_dispatch.py	161	def test_grad_through_dataarray_maximum():	CODE
48 more matches not shown…

Unused Imports65 hits · 62 pts

Severity	File	Line	Context
LOW	autograd/extend.py	2	CODE
LOW	autograd/extend.py	2	CODE
LOW	autograd/extend.py	2	CODE
LOW	autograd/extend.py	2	CODE
LOW	autograd/extend.py	2	CODE
LOW	autograd/extend.py	2	CODE
LOW	autograd/extend.py	2	CODE
LOW	autograd/extend.py	2	CODE
LOW	autograd/extend.py	2	CODE
LOW	autograd/extend.py	2	CODE
LOW	autograd/extend.py	2	CODE
LOW	autograd/extend.py	2	CODE
LOW	autograd/extend.py	16	CODE
LOW	autograd/extend.py	16	CODE
LOW	autograd/extend.py	16	CODE
LOW	autograd/extend.py	16	CODE
LOW	autograd/__init__.py	1	CODE
LOW	autograd/__init__.py	3	CODE
LOW	autograd/__init__.py	3	CODE
LOW	autograd/__init__.py	3	CODE
LOW	autograd/__init__.py	3	CODE
LOW	autograd/__init__.py	3	CODE
LOW	autograd/__init__.py	4	CODE
LOW	autograd/__init__.py	4	CODE
LOW	autograd/__init__.py	4	CODE
LOW	autograd/__init__.py	4	CODE
LOW	autograd/__init__.py	4	CODE
LOW	autograd/__init__.py	4	CODE
LOW	autograd/__init__.py	4	CODE
LOW	autograd/__init__.py	4	CODE
LOW	autograd/__init__.py	4	CODE
LOW	autograd/__init__.py	4	CODE
LOW	autograd/__init__.py	4	CODE
LOW	autograd/__init__.py	4	CODE
LOW	autograd/__init__.py	4	CODE
LOW	autograd/__init__.py	4	CODE
LOW	autograd/__init__.py	4	CODE
LOW	autograd/__init__.py	4	CODE
LOW	autograd/__init__.py	4	CODE
LOW	autograd/__init__.py	4	CODE
LOW	autograd/__init__.py	4	CODE
LOW	autograd/tracer.py	9	CODE
LOW	autograd/misc/__init__.py	1	CODE
LOW	autograd/misc/__init__.py	2	CODE
LOW	autograd/numpy/__init__.py	6	CODE
LOW	autograd/numpy/__init__.py	6	CODE
LOW	autograd/numpy/__init__.py	6	CODE
LOW	autograd/numpy/__init__.py	6	CODE
LOW	autograd/numpy/__init__.py	6	CODE
LOW	autograd/numpy/__init__.py	6	CODE
LOW	autograd/numpy/__init__.py	6	CODE
LOW	autograd/numpy/__init__.py	7	CODE
LOW	autograd/numpy/__init__.py	8	CODE
LOW	autograd/scipy/__init__.py	1	CODE
LOW	autograd/scipy/__init__.py	1	CODE
LOW	autograd/scipy/__init__.py	1	CODE
LOW	autograd/scipy/__init__.py	1	CODE
LOW	autograd/scipy/stats/__init__.py	1	CODE
LOW	autograd/scipy/stats/__init__.py	1	CODE
LOW	autograd/scipy/stats/__init__.py	1	CODE
5 more matches not shown…

Cross-Language Confusion2 hits · 10 pts

Severity	File	Line	Snippet	Context
HIGH	autograd/scipy/signal.py	13	f"Mode {mode} undefined, it can be one of 'valid', 'full', and 'same'"	CODE
HIGH	autograd/scipy/signal.py	153	f"Mode {mode} undefined, it can be one of 'valid', 'full', and 'same'"	CODE

Self-Referential Comments3 hits · 9 pts

Severity	File	Line	Snippet	Context
MEDIUM	tests/test_truediv.py	1	# This file is to check that future division works.	COMMENT
MEDIUM	examples/gaussian_process.py	41	# Define an example covariance function.	COMMENT
MEDIUM	examples/natural_gradient_black_box_svi.py	28	# Define the natural gradient	COMMENT

Fake / Example Data7 hits · 8 pts

Severity	File	Line	Snippet	Context
LOW	examples/variational_autoencoder.py	152	fake_data = generate_from_prior(gen_params, 20, latent_dim, seed)	CODE
LOW	examples/variational_autoencoder.py	153	save_images(fake_data, "vae_samples.png", vmin=0, vmax=1)	CODE
LOW	examples/generative_adversarial_net.py	65	fake_data = generate_from_noise(gen_params, num_samples, noise_dim, rs)	CODE
LOW	examples/generative_adversarial_net.py	66	logprobs_fake = logsigmoid(neural_net_predict(dsc_params, fake_data))	CODE
LOW⚡	examples/generative_adversarial_net.py	162	fake_data = generate_from_noise(gen_params, 20, noise_dim, seed)	CODE
LOW⚡	examples/generative_adversarial_net.py	164	probs_fake = np.mean(sigmoid(neural_net_predict(dsc_params, fake_data)))	CODE
LOW⚡	examples/generative_adversarial_net.py	167	save_images(fake_data, "gan_samples.png", vmin=0, vmax=1)	CODE

Over-Commented Block8 hits · 8 pts

Severity	File	Line	Snippet	Context
LOW	autograd/tracer.py	61	# order is determined by the dispatch mechanism. Also, if no other array-like	COMMENT
LOW	tests/test_graphs.py	201		COMMENT
LOW	tests/test_performance.py	1	# TODO:	COMMENT
LOW	tests/test_numpy.py	901		COMMENT
LOW	examples/define_gradient.py	21		COMMENT
LOW	examples/natural_gradient_black_box_svi.py	21	mu_density = np.sum(norm.logpdf(Y, mu, np.exp(log_sigma)), axis=1)	COMMENT
LOW	examples/gplvm.py	1	# Implements a Gaussian process latent-variable model.	COMMENT
LOW	.github/workflows/test.yml	41	allow-prereleases: true	COMMENT

AI Response Leakage1 hit · 8 pts

Severity	File	Line	Snippet	Context
HIGH	examples/gplvm.py	6	# In this example, we optimize X and the hyperparameters of the GP, but	COMMENT

Deep Nesting6 hits · 6 pts

Severity	File	Line	Context
LOW	autograd/tracer.py	118	CODE
LOW	autograd/numpy/numpy_wrapper.py	25	CODE
LOW	autograd/numpy/linalg.py	284	CODE
LOW	autograd/numpy/linalg.py	285	CODE
LOW	autograd/numpy/numpy_jvps.py	208	CODE
LOW	autograd/scipy/special.py	137	CODE

AI Slop Vocabulary3 hits · 6 pts

Severity	File	Line	Snippet	Context
LOW	autograd/numpy/linalg.py	263	# dimensions, so we just call a generic LU solve instead of directly using	COMMENT
LOW	autograd/numpy/numpy_boxes.py	20	# Basic array attributes just pass through	COMMENT
MEDIUM	examples/natural_gradient_black_box_svi.py	46	# is small (which leads to more robust/less chaotic ascent).	COMMENT

Excessive Try-Catch Wrapping1 hit · 1 pts

Severity	File	Line	Snippet	Context
LOW	autograd/tracer.py	88	except Exception as e:	CODE

Overly Generic Function Names1 hit · 1 pts

Severity	File	Line	Snippet	Context
LOW	autograd/numpy/numpy_vjps.py	286	def helper(g, n):	CODE

Analysis Overview

What These Metrics Mean

Score History

Severity Breakdown

Directory Score Breakdown

Pattern Findings