facebookresearch/faiss

4.4

Adjusted Score

4.4

Raw Score

100%

Time Factor

2026-07-13

Last Push

40.5K

Stars

C++

Language

243.5K

Lines of Code

916

Files

929

Pattern Hits

2026-07-14

Scan Date

0.02

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 16MEDIUM 36LOW 877

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 929 distinct pattern matches across 18 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 Identifiers376 hits · 376 pts

Severity	File	Line	Snippet	Context
LOW	tests/test_local_search_quantizer.py	187	def test_update_codebooks_with_double(self):	CODE
LOW	tests/test_local_search_quantizer.py	210	def test_compute_binary_terms(self):	CODE
LOW	tests/test_local_search_quantizer.py	472	def test_index_accuracy_reconstruct(self):	CODE
LOW	tests/test_local_search_quantizer.py	478	def test_index_accuracy_cqint(self):	CODE
LOW	tests/test_fast_scan_ivf.py	217	def test_by_residual_no_precomputed(self):	CODE
LOW	tests/test_fast_scan_ivf.py	268	def test_qloss_no_residual_ip(self):	CODE
LOW	tests/test_fast_scan_ivf.py	271	def test_qloss_by_residual_ip(self):	CODE
LOW	tests/test_fast_scan_ivf.py	327	def test_equiv_pq_encode_decode(self):	CODE
LOW⚡	tests/test_fast_scan_ivf.py	413	def test_no_residual_single_query(self):	CODE
LOW⚡	tests/test_fast_scan_ivf.py	416	def test_by_residual_single_query(self):	CODE
LOW⚡	tests/test_fast_scan_ivf.py	419	def test_no_residual_ip_single_query(self):	CODE
LOW⚡	tests/test_fast_scan_ivf.py	422	def test_by_residual_ip_single_query(self):	CODE
LOW⚡	tests/test_fast_scan_ivf.py	425	def test_no_residual_odd_dim_single_query(self):	CODE
LOW⚡	tests/test_fast_scan_ivf.py	428	def test_by_residual_odd_dim_single_query(self):	CODE
LOW	tests/test_fast_scan_ivf.py	1069	def test_fastscan_defaults_roundtrip(self):	CODE
LOW	tests/test_fast_scan_ivf.py	1082	def test_fastscan_max_lists_num_caps_probes(self):	CODE
LOW	tests/test_fast_scan_ivf.py	1101	def test_fastscan_ensure_topk_full_fills_heap(self):	CODE
LOW	tests/test_fast_scan_ivf.py	1119	def test_fastscan_ensure_topk_full_with_restrictive_selector(self):	CODE
LOW	tests/test_fast_scan_ivf.py	1140	def test_fastscan_ensure_topk_full_multi_query_works(self):	CODE
LOW	tests/test_fast_scan_ivf.py	1165	def test_fastscan_range_honors_max_empty_result_buckets(self):	CODE
LOW	tests/test_fast_scan_ivf.py	1201	def test_fastscan_fields_exposed(self):	CODE
LOW	tests/test_contrib.py	309	def test_equal_time_better_perf_shadows(self):	CODE
LOW	tests/test_contrib.py	328	def test_index_pretransformed(self):	CODE
LOW	tests/test_contrib.py	479	def test_range_search_preassigned_coarse_dis_none_matches_zeros(self):	CODE
LOW	tests/test_contrib.py	856	def test_ondisk_merge_with_shift_ids(self):	CODE
LOW	tests/test_contrib.py	874	def test_ivf_scalar_quantizer_types(self):	CODE
LOW	tests/test_contrib.py	905	def test_scalar_quantizer_types(self):	CODE
LOW	tests/test_contrib.py	933	def test_get_code_size_hnsw_non_default_m(self):	CODE
LOW	tests/test_contrib.py	944	def test_get_code_size_ivf_hnsw_non_default_m(self):	CODE
LOW	tests/test_contrib.py	954	def test_get_code_size_hnsw_roundtrip(self):	CODE
LOW	tests/test_graph_based.py	63	def test_hnsw_unbounded_queue(self):	CODE
LOW	tests/test_graph_based.py	346	def test_add_nan_vector_to_ivf_hnsw(self):	CODE
LOW	tests/test_graph_based.py	564	def test_nsg_supports_pre_built_knn_graph(self):	CODE
LOW	tests/test_graph_based.py	573	def test_nsg_with_pre_built_knn_graph_throws_when_rebuilding_via_add(self):	CODE
LOW	tests/test_graph_based.py	592	def test_nsg_rebuild_throws_with_pre_built_knn_graph(self):	CODE
LOW	tests/test_clustering_initialization.py	23	def verify_centroids_from_dataset(test_case, centroids, dataset):	CODE
LOW	tests/test_clustering_initialization.py	63	def test_initialization_methods(self):	CODE
LOW	tests/test_clustering_initialization.py	128	def test_with_existing_centroids(self):	CODE
LOW	tests/test_super_kmeans.py	14	def test_train_rejects_too_few_points(self):	CODE
LOW	tests/test_super_kmeans.py	20	def test_objective_decreases_monotonically(self):	CODE
LOW	tests/test_super_kmeans.py	35	def test_pruned_iter_assignments_close_to_vanilla(self):	CODE
LOW	tests/test_super_kmeans.py	56	def test_pruning_rate_in_expected_range(self):	CODE
LOW	tests/test_super_kmeans.py	77	def test_pdx_offset_correct_for_unaligned_y_batch(self):	CODE
LOW	tests/test_super_kmeans.py	132	def test_cluster_splitting_with_many_duplicates(self):	CODE
LOW	tests/test_super_kmeans.py	159	def test_train_populates_kmeans_surface(self):	CODE
LOW	tests/test_super_kmeans.py	177	def test_final_obj_matches_vanilla_kmeans(self):	CODE
LOW⚡	tests/test_svs_py.py	450	def test_svs_range_search_selected(self):	CODE
LOW⚡	tests/test_svs_py.py	454	def test_svs_add_search_remove_interface(self):	CODE
LOW⚡	tests/test_svs_py.py	458	def test_svs_batch_operations(self):	CODE
LOW⚡	tests/test_svs_py.py	635	def test_svs_leanvec_ood_training(self):	CODE
LOW⚡	tests/test_svs_py.py	640	def test_svs_leanvec_ood_training_smaller(self):	CODE
LOW⚡	tests/test_svs_py.py	643	def test_svs_leanvec_ood_training_wrong_dim(self):	CODE
LOW⚡	tests/test_svs_py.py	648	def test_svs_leanvec_ood_training_wrong_type(self):	CODE
LOW⚡	tests/test_svs_py.py	993	def test_ivf_leanvec_ood_training(self):	CODE
LOW⚡	tests/test_svs_py.py	1000	def test_ivf_leanvec_ood_training_smaller_queries(self):	CODE
LOW⚡	tests/test_svs_py.py	1006	def test_ivf_leanvec_ood_training_wrong_dim(self):	CODE
LOW⚡	tests/test_svs_py.py	1013	def test_ivf_leanvec_ood_training_wrong_type(self):	CODE
LOW⚡	tests/test_svs_py.py	1019	def test_ivf_leanvec_ood_search(self):	CODE
LOW⚡	tests/test_svs_py.py	1167	def test_static_ivf_add_throws(self):	CODE
LOW⚡	tests/test_svs_py.py	1175	def test_static_ivf_remove_throws(self):	CODE
316 more matches not shown…

Over-Commented Block240 hits · 239 pts

Severity	File	Line	Snippet	Context
LOW	CMakeLists.txt	1	# @lint-ignore-every LICENSELINT	COMMENT
LOW	pyproject-gpu-cuvs.toml	121	# manylinux_2_28 ships gcc-toolset-14; CUDA 13.2 supports GCC 14, so install	COMMENT
LOW	tutorial/python/11-SVS.py	1	# Portions Copyright (c) Meta Platforms, Inc. and affiliates.	COMMENT
LOW	tests/test_fastscan_filter.cpp	1	/*	COMMENT
LOW	tests/test_svs.cpp	21	* limitations under the License.	COMMENT
LOW	tests/test_hnsw.cpp	1	/*	COMMENT
LOW	tests/test_transfer_invlists.cpp	1	/*	COMMENT
LOW	tests/test_svs_py.py	1	# Portions Copyright (c) Meta Platforms, Inc. and affiliates.	COMMENT
LOW	tests/test_ivf_index.cpp	1	/*	COMMENT
LOW	tests/test_distances_dispatch.cpp	1	/*	COMMENT
LOW	tests/test_distances_dispatch.cpp	21	// * Functions returning an integer (fvec_L2sqr_ny_nearest's chosen	COMMENT
LOW	tests/test_ivf_early_termination.cpp	1	/*	COMMENT
LOW	tests/test_omp_exception_safety.cpp	1	/*	COMMENT
LOW	tests/test_omp_exception_safety.cpp	381	std::vector<idx_t> labels(1);	COMMENT
LOW	tests/test_mmap.cpp	1	/*	COMMENT
LOW	tests/test_scalar_quantizer.cpp	1	/*	COMMENT
LOW	tests/test_custom_result_handler.cpp	1	/*	COMMENT
LOW	tests/test_read_index_deserialize.cpp	1	/*	COMMENT
LOW	tests/test_read_index_deserialize.cpp	21	#include <faiss/IndexBinaryHNSW.h>	COMMENT
LOW	tests/test_read_index_deserialize.cpp	141	// before calling set_derived_values().	COMMENT
LOW	tests/test_read_index_deserialize.cpp	181	// Test: AdditiveQuantizer with nbits.size() != M causes out-of-bounds	COMMENT
LOW	tests/test_read_index_deserialize.cpp	441		COMMENT
LOW	tests/test_read_index_deserialize.cpp	821	push_val<int>(buf, 17); // b = 17 (exceeds 16 bits)	COMMENT
LOW	tests/test_read_index_deserialize.cpp	3661	// But include a few just in case:	COMMENT
LOW	.github/actions/build_cmake/action.yml	101		COMMENT
LOW	faiss/index_factory.cpp	21	#include <faiss/Index2Layer.h>	COMMENT
LOW	faiss/index_factory.cpp	41	#include <faiss/IndexNSG.h>	COMMENT
LOW	faiss/IndexFastScan.cpp	1	/*	COMMENT
LOW	faiss/factory_tools.cpp	21	#include <faiss/IndexIVFFlat.h>	COMMENT
LOW	faiss/CMakeLists.txt	601	endif()	COMMENT
LOW	faiss/IndexHNSW.h	41	// When set to false, level 0 in the knn graph is not initialized.	COMMENT
LOW	faiss/IndexHNSW.cpp	1	/*	COMMENT
LOW	faiss/IndexHNSW.cpp	21		COMMENT
LOW	faiss/IndexFlat.cpp	1	/*	COMMENT
LOW	faiss/AutoTune.cpp	21		COMMENT
LOW	faiss/IndexIVFPQ.cpp	21	#include <faiss/utils/distances_dispatch.h>	COMMENT
LOW	faiss/index_io.h	121	// When nonzero, deserialization rejects loop-driving fields (nlist,	COMMENT
LOW	faiss/index_io.h	141	// When nonzero, deserialization rejects IndexLattice payloads whose	COMMENT
LOW	faiss/IndexBinaryIVF.cpp	21	#include <faiss/IndexLSH.h>	COMMENT
LOW	faiss/SuperKMeans.h	1	/*	COMMENT
LOW	faiss/IndexIVFFlatPanorama.h	21	/// IDEA:	COMMENT
LOW	faiss/IndexIVFRaBitQFastScan.cpp	1	/*	COMMENT
LOW	faiss/SuperKMeans.cpp	1	/*	COMMENT
LOW	faiss/SuperKMeans.cpp	21	#include <faiss/impl/AdSampling.h>	COMMENT
LOW	faiss/IVFlib.cpp	1	/*	COMMENT
LOW	faiss/IndexIVFRaBitQFastScan.h	1	/*	COMMENT
LOW	faiss/IndexBinaryHNSW.cpp	21	#include <faiss/impl/AuxIndexStructures.h>	COMMENT
LOW	faiss/Clustering.h	41	/// If fewer than this number of training vectors per centroid are provided,	COMMENT
LOW	faiss/Clustering.h	61		COMMENT
LOW	faiss/IndexRowwiseMinMax.h	21	/// Basically, this index performs a rowwise scaling to [0,1] of every row	COMMENT
LOW	faiss/clone_index.cpp	21	#include <faiss/IndexEDEN.h>	COMMENT
LOW	faiss/clone_index.cpp	41	#include <faiss/IndexRefine.h>	COMMENT
LOW	faiss/impl/RaBitQuantizer.h	21	// with a Theoretical Error Bound for Approximate Nearest Neighbor Search".	COMMENT
LOW	faiss/impl/RaBitQuantizer.h	101	// abstraction layer due to its unique split encoding strategy	COMMENT
LOW	faiss/impl/Panorama.h	21	#include <cstring>	COMMENT
LOW	faiss/impl/Panorama.h	281	/// Panorama's core progressive filtering algorithm:	COMMENT
LOW	faiss/impl/RaBitQuantizer.cpp	1	/*	COMMENT
LOW	faiss/impl/RaBitQuantizerMultiBit.h	1	/*	COMMENT
LOW	faiss/impl/EDENQuantizer.h	61	}	COMMENT
LOW	faiss/impl/RaBitQUtils.h	1	/*	COMMENT
180 more matches not shown…

Unused Imports101 hits · 94 pts

Severity	File	Line	Context
LOW	tests/test_referenced_objects.py	7	CODE
LOW	tests/test_referenced_objects.py	7	CODE
LOW	tests/test_referenced_objects.py	7	CODE
LOW	tests/test_referenced_objects.py	7	CODE
LOW	tests/test_product_quantizer.py	6	CODE
LOW	tests/test_product_quantizer.py	6	CODE
LOW	tests/test_product_quantizer.py	6	CODE
LOW	tests/test_index_binary_from_float.py	6	CODE
LOW	tests/test_index_binary_from_float.py	6	CODE
LOW	tests/test_index_binary_from_float.py	6	CODE
LOW	tests/test_binary_io.py	8	CODE
LOW	tests/test_binary_io.py	8	CODE
LOW	tests/test_binary_io.py	8	CODE
LOW	tests/test_binary_io.py	8	CODE
LOW	tests/test_scalar_quantizer_correctness.py	6	CODE
LOW	tests/test_scalar_quantizer_correctness.py	6	CODE
LOW	tests/test_scalar_quantizer_correctness.py	6	CODE
LOW	tests/test_wheel_smoke.py	174	CODE
LOW	tests/bench_transposed_encode.py	6	CODE
LOW	tests/test_build_blocks.py	6	CODE
LOW	tests/test_build_blocks.py	6	CODE
LOW	tests/test_build_blocks.py	6	CODE
LOW	tests/test_binary_factory.py	6	CODE
LOW	tests/test_binary_factory.py	6	CODE
LOW	tests/test_binary_factory.py	6	CODE
LOW	tests/test_index.py	7	CODE
LOW	tests/test_index.py	7	CODE
LOW	tests/test_index.py	7	CODE
LOW	tests/test_omp_threads_py.py	6	CODE
LOW	tests/test_omp_threads_py.py	6	CODE
LOW	tests/test_omp_threads_py.py	6	CODE
LOW	tests/test_omp_threads_py.py	6	CODE
LOW	tests/test_index_accuracy.py	6	CODE
LOW	tests/test_index_accuracy.py	6	CODE
LOW	tests/test_index_accuracy.py	6	CODE
LOW	tests/test_ivflib.py	6	CODE
LOW	tests/test_ivflib.py	6	CODE
LOW	tests/test_ivflib.py	6	CODE
LOW	tests/common_faiss_tests.py	8	CODE
LOW	tests/common_faiss_tests.py	8	CODE
LOW	tests/common_faiss_tests.py	8	CODE
LOW	tests/common_faiss_tests.py	8	CODE
LOW	tests/test_standalone_codec.py	8	CODE
LOW	tests/test_clustering.py	6	CODE
LOW	tests/test_clustering.py	6	CODE
LOW	tests/test_clustering.py	6	CODE
LOW	contrib/torch/clustering.py	14	CODE
LOW	faiss/gpu/test/test_gpu_basics.py	6	CODE
LOW	faiss/gpu/test/test_gpu_basics.py	6	CODE
LOW	faiss/gpu/test/test_gpu_basics.py	6	CODE
LOW	faiss/gpu/test/test_gpu_basics.py	6	CODE
LOW	faiss/gpu/test/test_gpu_index_refs.py	6	CODE
LOW	faiss/gpu/test/test_gpu_index_ivfsq.py	6	CODE
LOW	faiss/gpu/test/test_gpu_index.py	6	CODE
LOW	faiss/gpu/test/test_gpu_index.py	6	CODE
LOW	faiss/gpu/test/test_gpu_index.py	6	CODE
LOW	faiss/gpu/test/test_gpu_index.py	6	CODE
LOW	faiss/python/extra_wrappers.py	13	CODE
LOW	faiss/python/__init__.py	128	CODE
LOW	faiss/python/__init__.py	132	CODE
41 more matches not shown…

Cross-File Repetition15 hits · 75 pts

Severity	File	Snippet	Context
HIGH	tests/test_ivf_flat_panorama.py	test correctness with various n_levels parameter values	STRING
HIGH	tests/test_flat_panorama.py	test correctness with various n_levels parameter values	STRING
HIGH	tests/test_refine_panorama.py	test correctness with various n_levels parameter values	STRING
HIGH	tests/test_ivf_flat_panorama.py	test edge case: more levels than dimension naturally supports	STRING
HIGH	tests/test_flat_panorama.py	test edge case: more levels than dimension naturally supports	STRING
HIGH	tests/test_refine_panorama.py	test edge case: more levels than dimension naturally supports	STRING
HIGH	tests/test_ivf_flat_panorama.py	test id selectors with dataset smaller than batch size	STRING
HIGH	tests/test_flat_panorama.py	test id selectors with dataset smaller than batch size	STRING
HIGH	tests/test_refine_panorama.py	test id selectors with dataset smaller than batch size	STRING
HIGH	tests/test_ivf_flat_panorama.py	test adding vectors incrementally in multiple batches	STRING
HIGH	tests/test_flat_panorama.py	test adding vectors incrementally in multiple batches	STRING
HIGH	tests/test_refine_panorama.py	test adding vectors incrementally in multiple batches	STRING
HIGH	tests/test_ivf_flat_panorama.py	test the correctness of the ratio of dimensions scanned	STRING
HIGH	tests/test_flat_panorama.py	test the correctness of the ratio of dimensions scanned	STRING
HIGH	tests/test_refine_panorama.py	test the correctness of the ratio of dimensions scanned	STRING

Deep Nesting78 hits · 74 pts

Severity	File	Line	Context
LOW	tests/test_local_search_quantizer.py	70	CODE
LOW	tests/test_graph_based.py	756	CODE
LOW	tests/test_rabitq_fastscan.py	180	CODE
LOW	tests/test_rabitq_fastscan.py	527	CODE
LOW	tests/test_rabitq_fastscan.py	675	CODE
LOW	tests/test_rabitq_fastscan.py	722	CODE
LOW	tests/test_rabitq_fastscan.py	774	CODE
LOW	tests/test_rabitq_fastscan.py	815	CODE
LOW	tests/test_rabitq_fastscan.py	853	CODE
LOW	tests/test_rabitq_fastscan.py	943	CODE
LOW	tests/test_ivf_flat_panorama.py	774	CODE
LOW	tests/test_flat_panorama.py	478	CODE
LOW	tests/test_flat_panorama.py	721	CODE
LOW	tests/test_scalar_quantizer_correctness.py	548	CODE
LOW	tests/test_fastscan_filter.py	31	CODE
LOW	tests/test_eden.py	207	CODE
LOW	tests/test_merge_index.py	23	CODE
LOW	tests/test_merge_index.py	91	CODE
LOW	tests/test_refine_panorama.py	470	CODE
LOW	tests/test_index_composite.py	720	CODE
LOW	tests/test_hnsw_panorama.py	353	CODE
LOW	tests/test_index.py	523	CODE
LOW	tests/test_rabitq.py	608	CODE
LOW	tests/test_rabitq.py	634	CODE
LOW	tests/test_rabitq.py	655	CODE
LOW	tests/test_rabitq.py	687	CODE
LOW	tests/test_rabitq.py	730	CODE
LOW	tests/test_rabitq.py	751	CODE
LOW	tests/test_rabitq.py	784	CODE
LOW	tests/test_rabitq.py	827	CODE
LOW	tests/test_rabitq.py	906	CODE
LOW	tests/test_index_accuracy.py	415	CODE
LOW	tests/test_refine.py	198	CODE
LOW	tests/test_search_params.py	26	CODE
LOW	contrib/exhaustive_search.py	60	CODE
LOW	contrib/datasets.py	493	CODE
LOW	contrib/factory_tools.py	10	CODE
LOW	contrib/factory_tools.py	83	CODE
LOW	contrib/big_batch_search.py	237	CODE
LOW	contrib/big_batch_search.py	219	CODE
LOW	faiss/gpu/test/bench_approaches.py	389	CODE
LOW	faiss/gpu/test/torch_test_contrib_gpu.py	288	CODE
LOW	faiss/gpu/test/torch_test_contrib_gpu.py	424	CODE
LOW	faiss/gpu/test/test_gpu_index_ivfsq.py	113	CODE
LOW	benchs/bench_hybrid_cpu_gpu.py	225	CODE
LOW	benchs/bench_hybrid_cpu_gpu.py	292	CODE
LOW	benchs/bench_quantizer.py	42	CODE
LOW	benchs/bench_pq_tables.py	19	CODE
LOW	benchs/bench_hnsw_prune_headroom.py	53	CODE
LOW	benchs/bench_gpu_1bn.py	705	CODE
LOW	benchs/bench_rabitq.py	156	CODE
LOW	benchs/bench_eden.py	89	CODE
LOW	benchs/bench_eden.py	129	CODE
LOW	benchs/bench_eden.py	351	CODE
LOW	benchs/bench_eden.py	497	CODE
LOW	benchs/bench_fw/optimize.py	130	CODE
LOW	benchs/bench_fw/benchmark.py	128	CODE
LOW	benchs/bench_fw/benchmark.py	1162	CODE
LOW	benchs/bench_fw/index.py	73	CODE
LOW	benchs/bench_fw/index.py	269	CODE
18 more matches not shown…

Decorative Section Separators14 hits · 45 pts

Severity	File	Line	Snippet	Context
MEDIUM	CMakeLists.txt	8	# =============================================================================	COMMENT
MEDIUM	CMakeLists.txt	20	# =============================================================================	COMMENT
MEDIUM⚡	tests/test_rabitq.py	485	# ==============================================================================	COMMENT
MEDIUM⚡	tests/test_rabitq.py	487	# ==============================================================================	COMMENT
MEDIUM	faiss/CMakeLists.txt	6	# =============================================================================	COMMENT
MEDIUM	faiss/CMakeLists.txt	10	# =============================================================================	COMMENT
MEDIUM	faiss/CMakeLists.txt	90	# =============================================================================	COMMENT
MEDIUM	faiss/CMakeLists.txt	528	# =============================================================================	COMMENT
MEDIUM	faiss/CMakeLists.txt	533	# =============================================================================	COMMENT
MEDIUM	faiss/gpu/CMakeLists.txt	7	# =============================================================================	COMMENT
MEDIUM	faiss/gpu/CMakeLists.txt	19	# =============================================================================	COMMENT
MEDIUM	faiss/gpu/test/CMakeLists.txt	7	# =============================================================================	COMMENT
MEDIUM	faiss/gpu/test/CMakeLists.txt	19	# =============================================================================	COMMENT
MEDIUM	faiss/utils/simd_impl/partitioning_simdlib256.h	47	// ── Per-TU SIMD type aliases ──────────────────────────────────────────	COMMENT

Structural Annotation Overuse26 hits · 44 pts

Severity	File	Line	Snippet	Context
LOW	INSTALL.md	118	## Step 1: invoking CMake	COMMENT
LOW	INSTALL.md	170	## Step 2: Invoking Make	COMMENT
LOW	INSTALL.md	205	## Step 3: Building the python bindings (optional)	COMMENT
LOW	INSTALL.md	216	## Step 4: Installing the C++ library and headers (optional)	COMMENT
LOW	INSTALL.md	227	## Step 5: Testing (optional)	COMMENT
LOW	.github/workflows/index-io-backward-compatibility.yml	23	# Step 1: Build with CMake and write files	COMMENT
LOW	.github/workflows/index-io-backward-compatibility.yml	48	# Step 2: Install conda faiss-cpu in a clean environment and read files	COMMENT
LOW	.github/workflows/index-io-backward-compatibility.yml	85	# Step 1: Install conda faiss-cpu in a clean environment and write files	COMMENT
LOW	.github/workflows/index-io-backward-compatibility.yml	114	# Step 2: Build with CMake and read files	COMMENT
LOW	faiss/impl/RaBitQuantizer.cpp	114	// Step 1: Compute 1-bit quantization and base factors	COMMENT
LOW	faiss/impl/RaBitQuantizer.cpp	154	// Step 2: Compute ex-bits quantization (if nb_bits > 1)	COMMENT
LOW	faiss/impl/RaBitQuantizerMultiBit.cpp	263	// Step 1: Compute L2 norm of residual	COMMENT
LOW⚡	faiss/impl/RaBitQuantizerMultiBit.cpp	278	// Step 2: Normalize residual	COMMENT
LOW⚡	faiss/impl/RaBitQuantizerMultiBit.cpp	284	// Step 3: Take absolute value	COMMENT
LOW⚡	faiss/impl/RaBitQuantizerMultiBit.cpp	290	// Step 4: Find optimal scaling factor	COMMENT
LOW⚡	faiss/impl/RaBitQuantizerMultiBit.cpp	293	// Step 5: Quantize to ex_bits	COMMENT
LOW⚡	faiss/impl/RaBitQuantizerMultiBit.cpp	304	// Step 6: Handle negative dimensions (flip bits)	COMMENT
LOW⚡	faiss/impl/RaBitQuantizerMultiBit.cpp	312	// Step 7: Pack codes into byte array	COMMENT
LOW	faiss/gpu_metal/MetalDistance.mm	273	// Step 1: coarse distance matrix	COMMENT
LOW	faiss/gpu_metal/MetalDistance.mm	297	// Step 2: coarse top-nprobe	COMMENT
LOW	faiss/gpu_metal/MetalDistance.mm	308	// Step 3: IVF scan	COMMENT
LOW	faiss/gpu_metal/MetalDistance.mm	340	// Step 4: merge	COMMENT
LOW	faiss/docs/simd_dynamic_dispatch_migration.md	97	### Step 1: Templatize on SIMDLevel	COMMENT
LOW	faiss/docs/simd_dynamic_dispatch_migration.md	138	### Step 2: Split into per-SIMD translation units	COMMENT
LOW	faiss/docs/simd_dynamic_dispatch_migration.md	195	### Step 3: Wire up dispatch	COMMENT
LOW	faiss/docs/simd_dynamic_dispatch_migration.md	245	### Step 4: Register in build system	COMMENT

Verbosity Indicators17 hits · 30 pts

Severity	File	Line	Snippet	Context
LOW	.github/workflows/index-io-backward-compatibility.yml	23	# Step 1: Build with CMake and write files	COMMENT
LOW	.github/workflows/index-io-backward-compatibility.yml	48	# Step 2: Install conda faiss-cpu in a clean environment and read files	COMMENT
LOW	.github/workflows/index-io-backward-compatibility.yml	85	# Step 1: Install conda faiss-cpu in a clean environment and write files	COMMENT
LOW	.github/workflows/index-io-backward-compatibility.yml	114	# Step 2: Build with CMake and read files	COMMENT
LOW	faiss/impl/RaBitQuantizer.cpp	114	// Step 1: Compute 1-bit quantization and base factors	COMMENT
LOW	faiss/impl/RaBitQuantizer.cpp	154	// Step 2: Compute ex-bits quantization (if nb_bits > 1)	COMMENT
LOW	faiss/impl/RaBitQuantizerMultiBit.cpp	263	// Step 1: Compute L2 norm of residual	COMMENT
LOW⚡	faiss/impl/RaBitQuantizerMultiBit.cpp	278	// Step 2: Normalize residual	COMMENT
LOW⚡	faiss/impl/RaBitQuantizerMultiBit.cpp	284	// Step 3: Take absolute value	COMMENT
LOW⚡	faiss/impl/RaBitQuantizerMultiBit.cpp	290	// Step 4: Find optimal scaling factor	COMMENT
LOW⚡	faiss/impl/RaBitQuantizerMultiBit.cpp	293	// Step 5: Quantize to ex_bits	COMMENT
LOW⚡	faiss/impl/RaBitQuantizerMultiBit.cpp	304	// Step 6: Handle negative dimensions (flip bits)	COMMENT
LOW⚡	faiss/impl/RaBitQuantizerMultiBit.cpp	312	// Step 7: Pack codes into byte array	COMMENT
LOW	faiss/gpu_metal/MetalDistance.mm	273	// Step 1: coarse distance matrix	COMMENT
LOW	faiss/gpu_metal/MetalDistance.mm	297	// Step 2: coarse top-nprobe	COMMENT
LOW	faiss/gpu_metal/MetalDistance.mm	308	// Step 3: IVF scan	COMMENT
LOW	faiss/gpu_metal/MetalDistance.mm	340	// Step 4: merge	COMMENT

AI Slop Vocabulary9 hits · 24 pts

Severity	File	Line	Snippet	Context
MEDIUM	tests/test_index_accuracy.py	252	index.nprobe = 4 # hopefully more robust than 1	CODE
MEDIUM	tests/test_index_accuracy.py	310	index.nprobe = 4 # hopefully more robust than 1	CODE
MEDIUM	faiss/IndexPQ.h	118	/// statistics are robust to internal threading, but not if	COMMENT
MEDIUM	faiss/IndexIVFPQ.h	175	/// statistics are robust to internal threading, but not if	COMMENT
MEDIUM	faiss/impl/PanoramaStats.h	17	/// Statistics are not robust to internal threading nor to	COMMENT
MEDIUM	faiss/gpu/test/test_gpu_index.py	264	# precomputed codes also utilize the coarse quantizer	COMMENT
LOW	faiss/python/array_conversions.py	90	# TODO: once deprecated classes are removed, remove the dict and just use .lower() below	COMMENT
MEDIUM	faiss/utils/hamming_distance/hamming_computer-avx512.h	16	// GenHammingComputer classes leverage SSE/AVX2 intrinsics.	COMMENT
MEDIUM	faiss/utils/hamming_distance/hamming_computer-avx2.h	14	// GenHammingComputer classes leverage SSE/AVX2 intrinsics.	COMMENT

Excessive Try-Catch Wrapping17 hits · 22 pts

Severity	File	Line	Snippet	Context
LOW	tests/test_io.py	574	except Exception:	CODE
LOW	tests/test_io.py	606	except Exception:	CODE
LOW	tests/test_build_blocks.py	219	except Exception:	CODE
LOW	…ndex_io_backward_compatibility/cmake_conda_io_utils.py	255	except Exception as e:	CODE
LOW	…ndex_io_backward_compatibility/cmake_conda_io_utils.py	288	except Exception as e:	CODE
LOW	…ndex_io_backward_compatibility/cmake_conda_io_utils.py	318	except Exception as e:	CODE
LOW	…ndex_io_backward_compatibility/cmake_conda_io_utils.py	348	except Exception as e:	CODE
LOW	contrib/rpc.py	145	except Exception as e:	CODE
LOW⚡	faiss/gpu/test/test_index_cpu_to_gpu.py	43	except Exception as e:	CODE
LOW⚡	faiss/gpu/test/test_index_cpu_to_gpu.py	59	except Exception as e:	CODE
LOW	faiss/gpu/test/test_index_cpu_to_gpu.py	97	except Exception as e:	CODE
MEDIUM	faiss/gpu/test/test_index_cpu_to_gpu.py	40	def verify_throws_not_implemented_exception(self, factory_string):	CODE
MEDIUM	faiss/gpu/test/test_index_cpu_to_gpu.py	50	def verify_clones_successfully(	CODE
MEDIUM	faiss/gpu/test/test_index_cpu_to_gpu.py	92	def test_with_flag_set_to_false(self):	CODE
LOW	benchs/bench_all_ivf/parse_bench_all_ivf.py	139	except Exception as e:	CODE
LOW	demos/offline_ivf/offline_ivf.py	615	except Exception:	CODE
MEDIUM	demos/offline_ivf/offline_ivf.py	603	def _knn_function(self, xq, xb, k, metric, thread_id=None):	CODE

Self-Referential Comments7 hits · 21 pts

Severity	File	Line	Snippet	Context
MEDIUM	tests/test_graph_based.py	357	# Create a vector with NaN in the first component	COMMENT
MEDIUM	tests/test_flat_panorama.py	686	# Create a random permutation	COMMENT
MEDIUM	tests/test_wheel_smoke_gpu.py	26	# This file is the faiss-gpu wheel smoke test. It also gets collected by	COMMENT
MEDIUM	tests/test_hnsw_panorama.py	433	# Create a permutation (reverse order for simplicity)	COMMENT
MEDIUM	tests/test_index.py	817	# Create the dataset	COMMENT
MEDIUM	tests/test_search_params.py	166	# Create a mask of valid IDs (those in subset)	COMMENT
MEDIUM	faiss/gpu/hipify.sh	42	# Create a backup of the original gpu directory	COMMENT

Modern Structural Boilerplate10 hits · 10 pts

Severity	File	Line	Snippet	Context
LOW	contrib/client_server.py	65	def set_nprobe(self, nprobe: int) -> None:	CODE
LOW	contrib/client_server.py	68	def set_omp_num_threads(self, nt: int) -> None:	CODE
LOW	faiss/python/__init__.py	166	logger = logging.getLogger(__name__)	CODE
LOW	faiss/python/loader.py	107	logger = logging.getLogger(__name__)	CODE
LOW	benchs/bench_fw/optimize.py	20	logger = logging.getLogger(__name__)	CODE
LOW	benchs/bench_fw/benchmark.py	32	logger = logging.getLogger(__name__)	CODE
LOW	benchs/bench_fw/index.py	39	logger = logging.getLogger(__name__)	CODE
LOW	benchs/bench_fw/benchmark_io.py	24	logger = logging.getLogger(__name__)	CODE
LOW	benchs/bench_fw/descriptors.py	16	logger = logging.getLogger(__name__)	CODE
LOW	benchs/bench_fw/utils.py	19	logger = logging.getLogger(__name__)	CODE

Redundant / Tautological Comments4 hits · 6 pts

Severity	File	Line	Snippet	Context
LOW	tests/test_svs_py.py	30	# Check if LVQ/LeanVec support is available	COMMENT
LOW	tests/test_hnsw_panorama.py	406	# Check if this neighbor is still in the result set	COMMENT
LOW	faiss/python/class_wrappers.py	260	# Check if the attribute already exists (valid attribute)	COMMENT
LOW	faiss/python/class_wrappers.py	262	# Check if it exists on the instance or class	COMMENT

AI Structural Patterns6 hits · 6 pts

Severity	File	Line	Context
LOW	contrib/big_batch_search.py	237	CODE
LOW	faiss/gpu/test/bench_approaches.py	258	CODE
LOW	benchs/kmeans_mnist.py	35	CODE
LOW	benchs/bench_hnsw_prune_headroom.py	53	CODE
LOW	benchs/bench_fw/utils.py	178	CODE
LOW	demos/offline_ivf/tests/testing_utils.py	29	CODE

Docstring Block Structure1 hit · 5 pts

Severity	File	Line	Snippet	Context
HIGH	contrib/vecs_io.py	85	Memory-mapped iterator over chunked .bvecs files. Iterates through all chunk files in order (chunk_0000.bvecs,	STRING

Example Usage Blocks3 hits · 4 pts

Severity	File	Line	Snippet	Context
LOW	faiss/impl/ClusteringInitialization.h	46	/// Example usage:	COMMENT
LOW	faiss/utils/extra_distances.h	63	/// Example usage:	COMMENT
LOW	demos/index_pq_flat_separate_codes_from_codebook.py	256	# # Example usage:	COMMENT

Slop Phrases3 hits · 4 pts

Severity	File	Line	Snippet	Context
LOW	tests/test_build_blocks.py	335	# make sure to test extreme cases	COMMENT
LOW	tests/test_index_accuracy.py	583	# make sure to test below and above	COMMENT
MEDIUM	contrib/rpc.py	10	This code is for demonstration purposes only, and does not include certain	STRING

Dead Code2 hits · 4 pts

Severity	File	Line	Snippet	Context
MEDIUM	benchs/bench_fw/index.py	149		CODE
MEDIUM	benchs/bench_fw/index.py	150		CODE

Analysis Overview

What These Metrics Mean

Score History

Severity Breakdown

Directory Score Breakdown

Pattern Findings