2.3

Adjusted Signal

2.3

Raw Pattern Score

100%

Time Factor

2026-10-31

Last Push

4.1K

Stars

C++

Language

1.7M

Lines of Code

4.4K

Files

4.2K

Pattern Hits

2026-10-31

Scan Date

0.00

HC Hit Rate

Risk Review · Payload v1

Use each vector to guide manual verification of the underlying findings.

Payload schema: 1.0.0 · Engine: 2.0.0 · Scoring: heuristic-density-v1 · Scan date: 2026-10-31.

Cognitive Ownership Risk 9.4/100 · LOW

Can maintainers explain intent, assumptions, and trade-offs? 3991 finding(s).

Hallucinated Dependency & Supply Chain Risk 0.0/100 · LOW

Are imports declared and package/API names verified? 1 finding(s).

Agent Over-Engineering Ratio 1.6/100 · LOW

Are wrappers, guards, and repeated structures necessary? 189 finding(s).

PR Review Velocity Impact 12.5/100 · LOW

Which changes need focused human review? 4167 finding(s).

What These Metrics Mean

Adjusted Signal: Legacy pattern score normalized per 1,000 scanned lines and multiplied by the repository-level time factor. It is not an estimate of AI authorship.
Raw Pattern Score: Sum of severity- and context-weighted pattern findings before applying the time factor.
Time Factor: A coarse multiplier based on repository last-push metadata. It does not identify when individual code was written.
Pattern Hits: Count of configured pattern findings. Review each location and surrounding code before acting.
HC Hit Rate: High- and critical-severity findings per scanned file. Severity is a configured detector weight, not proof of a vulnerability or authorship.
Lines / Files: Scan coverage denominators used by the legacy normalized score.

Pattern-Signal History

This chart tracks the legacy adjusted pattern score across scan runs. Changes can reflect source changes, scan coverage, or detector/scoring versions; they do not measure AI adoption or prove that a team refactored code. Compare runs only when their scoring versions are compatible. The dashed line tracks the raw pattern-hit count.

Severity Breakdown

Groups findings by configured severity weight. Severity is not a confidence estimate, security rating, or proof that a package is missing. Verify imports against the project's manifests and inspect each finding in its local design context.

CRITICAL 0HIGH 5MEDIUM 205LOW 3957

Directory Score Breakdown

This horizontal bar chart shows where configured pattern scores accumulate across directories. Treat the bars as locations to inspect, not as evidence of authorship or code quality. Use the file-level findings to decide whether a directory needs a focused ownership, dependency, complexity, or review check.

Pattern Findings

The scanner identified 4167 distinct pattern matches across 17 syntactic categories. Each entry below represents a discrete location in the source code where the engine recorded a detector match to inspect in context. 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 configured finding weight (CRITICAL / HIGH / MEDIUM / LOW). The Context column identifies whether the match occurred inside executable code, an inline comment, or a string literal — legacy comment-context matches receive a ×1.5 score weight. The ⚡ bolt icon marks clustered matches: three or more patterns within a 10-line window, each receiving an additional ×1.5 density multiplier under the legacy scoring model.

Over-Commented Block3831 hits · 3108 pts

Severity	File	Line	Snippet	Context
LOW	CMakeLists.txt	1	#===============================================================================	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	21		COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	101	/// @addtogroup dnnl_api_cpp_primitives Primitives	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	161	primitive() = default;	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	181	///	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	201	/// @note The cache blob can be empty. It's the user's responsibility to	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	261		COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	281	/// scratchpad is common to all primitives to reduce the memory footprint.	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	301	///	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	321	inline dnnl_rounding_mode_t convert_to_c(rounding_mode mode) {	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	341	/// static quantization mode with scale applied before zero-point:	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	361	forward_training = dnnl_forward_training,	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	461	lrn_within_channel = dnnl_lrn_within_channel,	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	541	/// @param aalgorithm C++ API algorithm kind enum value.	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	561	/// the same behavior as #dnnl::prop_kind::backward.	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	581	use_shift = dnnl_use_shift,	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	601	/// On backward propagation, the result of the backward ReLU operation	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	661	bidirectional_sum = dnnl_bidirectional_sum,	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	821		COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	841	/// The latter approach requires understanding how the physical data	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	861	/// when used, can be queried using #dnnl::sycl_interop::get_buffer and set	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	881	/// library because there is no clear definition of what the output values	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	921	/// @deprecated The f4_e3m0 data type will be removed in a future release.	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	981	opaque = dnnl_format_kind_opaque,	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	1001	grouped = dnnl_grouped,	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	1021	/// - Domain-specific names, i.e. names that make sense only in the	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	1041	/// - @c 'd', @c 'h', and @c 'w' denote spatial depth, height, and width	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	1221	hwigo = decab,	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	1241	ldnc = abcd,	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	1261	ldio = abcd,	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	2881	reset(zero_md);	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	2941	}	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	2981	/// contains n+1 buffers for an n-dimensional tensor.	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	3021	/// is created using the queried memory descriptor are unspecified	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	3041	"could not create a memory descriptor for packed "	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	3061	/// - Variable dimension: 0 (token count varies per expert)	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	3121	/// Construct a memory descriptor from a binary blob.	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	3161	/// The operation ensures that the transformation of the physical memory	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	3181	/// their logical counterparts. This also assumes that these	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	3201	dnnl_memory_desc_t out_md = nullptr;	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	3221	/// #dnnl::memory::format_kind::any.	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	3281	memory::dim get_submemory_offset() const {	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	3301	/// @note	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	3321	}	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	3481	/// will be initialized as if #dnnl::memory::set_data_handle() had been	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	3501	/// Unless @p handle is equal to #DNNL_MEMORY_NONE, the constructed memory	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	3541		COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	3661	///	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	3681	template <typename T = void>	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	3741		COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	3801	/// The kind of this post-op is #dnnl::primitive::kind::sum.	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	3861	/// @param zero_point Single scalar int32_t value of zeropoint.	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	3901	"could not get parameters of an elementwise post-op");	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	3981	/// dst[:] <- binary_op (dst[:], another_input[:])	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	4001	/// dst[:] <- binary_op (dst[:], another_input1[:], another_input2[:])	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	4061	error::wrap_c_api(dnnl_memory_desc_clone(&cloned_md2, cdesc2),	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	4081	/// oh = 14, ow = 14; // convolution output params	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	4101	/// @note	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	4341	get(), dnnl::convert_to_c(mode)),	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	4361	}	COMMENT
3771 more matches not shown…

Decorative Section Separators169 hits · 514 pts

Severity	File	Line	Snippet	Context
MEDIUM	CMakeLists.txt	1	#===============================================================================	COMMENT
MEDIUM	CMakeLists.txt	16	#===============================================================================	COMMENT
MEDIUM	tests/CMakeLists.txt	1	#===============================================================================	COMMENT
MEDIUM	tests/CMakeLists.txt	15	#===============================================================================	COMMENT
MEDIUM	tests/generate_c_symbols_refs.sh	2	#===============================================================================	COMMENT
MEDIUM	tests/generate_c_symbols_refs.sh	16	#===============================================================================	COMMENT
MEDIUM	tests/benchdnn/CMakeLists.txt	1	#===============================================================================	COMMENT
MEDIUM	tests/benchdnn/CMakeLists.txt	15	#===============================================================================	COMMENT
MEDIUM	tests/other/subproject/CMakeLists.txt	1	#===============================================================================	COMMENT
MEDIUM	tests/other/subproject/CMakeLists.txt	15	#===============================================================================	COMMENT
MEDIUM	tests/noexcept/CMakeLists.txt	1	#===============================================================================	COMMENT
MEDIUM	tests/noexcept/CMakeLists.txt	15	#===============================================================================	COMMENT
MEDIUM	tests/gtests/CMakeLists.txt	1	#===============================================================================	COMMENT
MEDIUM	tests/gtests/CMakeLists.txt	15	#===============================================================================	COMMENT
MEDIUM	tests/gtests/ze/CMakeLists.txt	1	#===============================================================================	COMMENT
MEDIUM	tests/gtests/ze/CMakeLists.txt	15	#===============================================================================	COMMENT
MEDIUM	tests/gtests/ze/api/CMakeLists.txt	1	#===============================================================================	COMMENT
MEDIUM	tests/gtests/ze/api/CMakeLists.txt	15	#===============================================================================	COMMENT
MEDIUM	tests/gtests/graph/CMakeLists.txt	1	#===============================================================================	COMMENT
MEDIUM	tests/gtests/graph/CMakeLists.txt	15	#===============================================================================	COMMENT
MEDIUM	tests/gtests/graph/unit/CMakeLists.txt	1	#===============================================================================	COMMENT
MEDIUM	tests/gtests/graph/unit/CMakeLists.txt	16	#===============================================================================	COMMENT
MEDIUM	tests/gtests/graph/unit/interface/CMakeLists.txt	1	#===============================================================================	COMMENT
MEDIUM	tests/gtests/graph/unit/interface/CMakeLists.txt	15	#===============================================================================	COMMENT
MEDIUM	tests/gtests/graph/unit/interface/sycl/CMakeLists.txt	1	#===============================================================================	COMMENT
MEDIUM	tests/gtests/graph/unit/interface/sycl/CMakeLists.txt	15	#===============================================================================	COMMENT
MEDIUM	tests/gtests/graph/unit/utils/CMakeLists.txt	1	#===============================================================================	COMMENT
MEDIUM	tests/gtests/graph/unit/utils/CMakeLists.txt	15	#===============================================================================	COMMENT
MEDIUM	tests/gtests/graph/unit/backend/CMakeLists.txt	1	#===============================================================================	COMMENT
MEDIUM	tests/gtests/graph/unit/backend/CMakeLists.txt	15	#===============================================================================	COMMENT
MEDIUM	tests/gtests/graph/unit/backend/dnnl/CMakeLists.txt	1	#===============================================================================	COMMENT
MEDIUM	tests/gtests/graph/unit/backend/dnnl/CMakeLists.txt	15	#===============================================================================	COMMENT
MEDIUM	tests/gtests/graph/unit/backend/fake/CMakeLists.txt	1	#===============================================================================	COMMENT
MEDIUM	tests/gtests/graph/unit/backend/fake/CMakeLists.txt	15	#===============================================================================	COMMENT
MEDIUM	tests/gtests/graph/api/CMakeLists.txt	1	#===============================================================================	COMMENT
MEDIUM	tests/gtests/graph/api/CMakeLists.txt	15	#===============================================================================	COMMENT
MEDIUM	tests/gtests/graph/api/sycl/CMakeLists.txt	1	#===============================================================================	COMMENT
MEDIUM	tests/gtests/graph/api/sycl/CMakeLists.txt	15	#===============================================================================	COMMENT
MEDIUM	tests/gtests/graph/api/ocl/CMakeLists.txt	1	#===============================================================================	COMMENT
MEDIUM	tests/gtests/graph/api/ocl/CMakeLists.txt	15	#===============================================================================	COMMENT
MEDIUM	tests/gtests/internals/CMakeLists.txt	1	#===============================================================================	COMMENT
MEDIUM	tests/gtests/internals/CMakeLists.txt	15	#===============================================================================	COMMENT
MEDIUM	tests/gtests/sycl/CMakeLists.txt	1	#===============================================================================	COMMENT
MEDIUM	tests/gtests/sycl/CMakeLists.txt	15	#===============================================================================	COMMENT
MEDIUM	tests/gtests/sycl/api/CMakeLists.txt	1	#===============================================================================	COMMENT
MEDIUM	tests/gtests/sycl/api/CMakeLists.txt	15	#===============================================================================	COMMENT
MEDIUM	tests/gtests/api/CMakeLists.txt	1	#===============================================================================	COMMENT
MEDIUM	tests/gtests/api/CMakeLists.txt	15	#===============================================================================	COMMENT
MEDIUM	tests/gtests/regression/CMakeLists.txt	1	#===============================================================================	COMMENT
MEDIUM	tests/gtests/regression/CMakeLists.txt	15	#===============================================================================	COMMENT
MEDIUM	tests/gtests/ocl/CMakeLists.txt	1	#===============================================================================	COMMENT
MEDIUM	tests/gtests/ocl/CMakeLists.txt	15	#===============================================================================	COMMENT
MEDIUM	tests/gtests/ocl/api/CMakeLists.txt	1	#===============================================================================	COMMENT
MEDIUM	tests/gtests/ocl/api/CMakeLists.txt	15	#===============================================================================	COMMENT
MEDIUM	third_party/kleidiai/experimental/ops/CMakeLists.txt	7	# -------------------------------------------------------------------------------------------------	COMMENT
MEDIUM	third_party/kleidiai/experimental/ops/CMakeLists.txt	27	# -------------------------------------------------------------------------------------------------	COMMENT
MEDIUM⚡	third_party/kleidiai/experimental/ops/CMakeLists.txt	75	# -------------------------------------------------------------------------------------------------	COMMENT
MEDIUM⚡	third_party/kleidiai/experimental/ops/CMakeLists.txt	80	# -------------------------------------------------------------------------------------------------	COMMENT
MEDIUM⚡	third_party/kleidiai/experimental/ops/CMakeLists.txt	84	# -------------------------------------------------------------------------------------------------	COMMENT
MEDIUM⚡	third_party/kleidiai/experimental/ops/CMakeLists.txt	90	# -------------------------------------------------------------------------------------------------	COMMENT
109 more matches not shown…

AI Slop Vocabulary34 hits · 102 pts

Severity	File	Line	Snippet	Context
MEDIUM	tests/benchdnn/dnnl_common.hpp	300	// All these memories must utilize `prefill_memory` flag of `dnn_mem_t` ctor	COMMENT
MEDIUM	tests/benchdnn/dnnl_common.cpp	81	// Service primitive is not supposed to utilize further logic.	COMMENT
MEDIUM	tests/benchdnn/dnnl_common.cpp	2142	// Rationale: make sure that the primitive cache is robust in the case	COMMENT
MEDIUM	tests/benchdnn/eltwise/eltwise.cpp	485	// Unique scenario when driver can utilize inplace but original source is	COMMENT
MEDIUM	tests/benchdnn/utils/res.hpp	92	// for a prim_ref (--fast-ref) test object which can utilize blocked	COMMENT
MEDIUM	tests/benchdnn/utils/compare.hpp	119	// they may be relevant in specific cases. This is a hint to utilize	COMMENT
MEDIUM	tests/benchdnn/matmul/matmul.cpp	607	// sparse problems. `CSR`/`COO` will utilize their `fill_sparse_data`	COMMENT
MEDIUM	tests/benchdnn/rnn/rnn.hpp	88	// We need a more robust way of testing RNN.	COMMENT
MEDIUM⚡	third_party/level_zero/level_zero/ze_api.h	7722	ZE_CACHE_EXT_REGION_ZE_CACHE_REGION_DEFAULT = 0, ///< [DEPRECATED] utilize driver default sch	CODE
MEDIUM⚡	third_party/level_zero/level_zero/ze_api.h	7724	ZE_CACHE_EXT_REGION_ZE_CACHE_RESERVE_REGION = 1, ///< [DEPRECATED] utilize reserved region. U	CODE
MEDIUM⚡	third_party/level_zero/level_zero/ze_api.h	7726	ZE_CACHE_EXT_REGION_ZE_CACHE_NON_RESERVED_REGION = 2, ///< [DEPRECATED] utilize non-reserverd regi	CODE
MEDIUM⚡	third_party/level_zero/level_zero/ze_api.h	7728	ZE_CACHE_EXT_REGION_DEFAULT = 0, ///< utilize driver default scheme	CODE
MEDIUM⚡	third_party/level_zero/level_zero/ze_api.h	7729	ZE_CACHE_EXT_REGION_RESERVED = 1, ///< utilize reserved region	CODE
MEDIUM⚡	third_party/level_zero/level_zero/ze_api.h	7730	ZE_CACHE_EXT_REGION_NON_RESERVED = 2, ///< utilize non-reserverd region	CODE
MEDIUM	third_party/spdlog/spdlog/fmt/bundled/format.h	3207	// Compilers are not aware of how to leverage the maximum value of	COMMENT
MEDIUM	third_party/gtest/include/gtest/gtest-death-test.h	151	// This implementation is not meant to be as highly tuned or robust	COMMENT
MEDIUM	third_party/gtest/include/gtest/gtest-death-test.h	222	// In practice, this can be used to test functions that utilize the	COMMENT
MEDIUM	examples/graph/sdpa_dynamic_quantized.cpp	277	// leverage the tensor API to allocate the scratchpad buffer for	COMMENT
MEDIUM	examples/graph/gqa.cpp	196	// leverage the tensor API to allocate the scratchpad buffer for	COMMENT
MEDIUM	src/gpu/intel/reorder/custom.cpp	387	// access mem using large enough chunks to utilize whole cache lines.	COMMENT
MEDIUM	…pu/intel/gemm/jit/generator/pieces/register_layout.cpp	237	// A64 Essentially max 256 bytes.	COMMENT
MEDIUM	src/gpu/intel/softmax/reusable.hpp	177	// utilize largest supported subgroup size	COMMENT
MEDIUM	src/cpu/x64/jit_avx512_common_1x1_convolution.hpp	188	// Note: A robust fusion implementation would be to check if both	COMMENT
MEDIUM	src/cpu/x64/jit_avx512_core_amx_1x1_conv_kernel.cpp	1211	// case if we utilize half of tile or less.	COMMENT
MEDIUM	src/cpu/x64/jit_avx2_1x1_convolution.hpp	195	// Note: A robust fusion implementation would be to check if both	COMMENT
MEDIUM	…c/cpu/x64/jit_avx512_core_x8s8s32x_1x1_convolution.hpp	201	// Note: A robust fusion implementation would be to check if both	COMMENT
MEDIUM	src/cpu/x64/jit_avx512_core_bf16_1x1_convolution.hpp	220	// Note: A robust fusion implementation would be to check if both	COMMENT
MEDIUM	src/cpu/x64/jit_uni_x8s8s32x_1x1_convolution.hpp	243	// Note: A robust fusion implementation would be to check if both	COMMENT
MEDIUM	src/cpu/x64/injectors/injector_utils.hpp	135	// This constructor is used to utilize extended registers (e.g. r16-r31).	COMMENT
MEDIUM	src/cpu/x64/brgemm/brgemm_types.hpp	770	// TODO: extend usage to primitives (each of them utilize their own copy	COMMENT
MEDIUM	src/cpu/x64/matmul/brgemm_matmul_utils.cpp	1170	// - M-parallelism is too low to fully utilize the cores.	COMMENT
MEDIUM	src/cpu/aarch64/jit_sve_1x1_convolution.hpp	236	// Note: A robust fusion implementation would be to check if both	COMMENT
MEDIUM	src/cpu/aarch64/rnn/rnn_brgemm_utils.cpp	226	// inherited 32 degrades badly at 32). 40 is the most robust single value across	COMMENT
MEDIUM	src/xpu/sycl/types.hpp	128	// would utilize it.	COMMENT

Verbosity Indicators34 hits · 61 pts

Severity	File	Line	Snippet	Context
LOW⚡	tests/benchdnn/sdpa/ref_sdpa.cpp	240	// Step 1: score = Q x K (matmul primitive).	COMMENT
LOW⚡	tests/benchdnn/sdpa/ref_sdpa.cpp	244	// Step 2: Scale.	COMMENT
LOW⚡	tests/benchdnn/sdpa/ref_sdpa.cpp	247	// Step 3: Apply attention mask (buffer or causal).	COMMENT
LOW	tests/benchdnn/sdpa/ref_sdpa.cpp	288	// Step 4: Softmax over K dimension (axis = 2 of the 3-D score tensor).	COMMENT
LOW	tests/benchdnn/sdpa/ref_sdpa.cpp	306	// Step 5: output = prob_dp x V (matmul primitive).	COMMENT
LOW	third_party/spdlog/spdlog/fmt/bundled/format-inl.h	1260	// Step 1: integer promotion & Schubfach multiplier calculation.	COMMENT
LOW	third_party/spdlog/spdlog/fmt/bundled/format-inl.h	1315	// Step 2: Try larger divisor; remove trailing zeros if necessary.	COMMENT
LOW	third_party/spdlog/spdlog/fmt/bundled/format-inl.h	1347	// Step 3: Find the significand with the smaller divisor.	COMMENT
LOW⚡	examples/matmul_f8_quantization.cpp	257	// Step 1: Compute scaling factors for quantization	COMMENT
LOW⚡	examples/matmul_f8_quantization.cpp	266	// Step 2: Quantize f32 to f8 format with scaling	COMMENT
LOW	examples/matmul_f8_quantization.cpp	340	// Step 3: Matrix multiplication with f8	COMMENT
LOW	examples/matmul_f8_quantization.cpp	383	// Step 4: Validate results	COMMENT
LOW	src/gpu/intel/reduction/utils.cpp	191	// Step 7: Potentially add dst-zero-padding if needed for the final reduction dimensions.	COMMENT
LOW	src/cpu/platform.cpp	324	// The purpose of this function is to return the potential maximum number of	COMMENT
LOW	src/cpu/platform.cpp	379	/* The purpose of this function is to provide a very efficient timestamp	COMMENT
LOW	src/cpu/rv64/rvv_winograd_convolution.cpp	54	// Step 1: temp = G * filter (4x3)	COMMENT
LOW	src/cpu/rv64/rvv_winograd_convolution.cpp	66	// Step 2: result = temp * G^T, store directly to GEMM layout	COMMENT
LOW	src/cpu/rv64/rvv_winograd_convolution.cpp	655	// Step 1: Input transform via JIT kernel (tile loop is inside JIT)	COMMENT
LOW	src/cpu/rv64/rvv_winograd_convolution.cpp	666	// Step 2: GEMM using brgemm kernel (16 elements)	COMMENT
LOW	src/cpu/rv64/rvv_winograd_convolution.cpp	677	// Step 3: Output transform via JIT kernel (tile loop is inside JIT)	COMMENT
LOW⚡	src/cpu/x64/ir/reg_alloc.cpp	593	// Step 1: compute operation-level liveness.	COMMENT
LOW⚡	src/cpu/x64/ir/reg_alloc.cpp	599	// Step 2: build a single live interval per virtual register.	COMMENT
LOW⚡	src/cpu/x64/ir/reg_alloc.cpp	638	// Step 3: weight each virtual register by how often it is referenced and	COMMENT
LOW⚡	src/cpu/x64/ir/reg_alloc.cpp	643	// Step 4: run linear-scan register allocation per physical register file,	COMMENT
LOW	src/cpu/x64/brgemm/brgemm_utils.hpp	61	/* The purpose of this function is to enable initialization of brgemm values	COMMENT
LOW	src/cpu/x64/brgemm/brgemm_utils.hpp	72	/* The purpose of this function is to enable initialization of brgemm values	COMMENT
LOW	src/cpu/aarch64/brgemm/brgemm_utils.hpp	119	/* The purpose of this function is to enable initialization of brgemm values	COMMENT
LOW	src/cpu/aarch64/brgemm/brgemm_utils.hpp	130	/* The purpose of this function is to enable initialization of brgemm values	COMMENT
LOW	src/graph/backend/dnnl/layout_propagator.cpp	1260	// if the output layout is specified, we need to check if it's matched	COMMENT
LOW	src/common/dnnl_thread.hpp	144	/* The purpose of this function is to provide the number of threads the library	COMMENT
LOW⚡	src/common/primitive_desc_iface.cpp	108	// Step 1: create impl::primitive_t or get it from primitive cache	COMMENT
LOW⚡	src/common/primitive_desc_iface.cpp	116	// Step 2: create primitive_iface_t, init and return it to user	COMMENT
LOW⚡	src/common/reorder_primitive_desc_iface.hpp	68	// Step 1: create impl::primitive_t or get it from primitive cache	COMMENT
LOW⚡	src/common/reorder_primitive_desc_iface.hpp	76	// Step 2: create primitive_iface_t, init and return it to user	COMMENT

Structural Annotation Overuse30 hits · 58 pts

Severity	File	Line	Snippet	Context
LOW⚡	tests/benchdnn/sdpa/ref_sdpa.cpp	240	// Step 1: score = Q x K (matmul primitive).	COMMENT
LOW⚡	tests/benchdnn/sdpa/ref_sdpa.cpp	244	// Step 2: Scale.	COMMENT
LOW⚡	tests/benchdnn/sdpa/ref_sdpa.cpp	247	// Step 3: Apply attention mask (buffer or causal).	COMMENT
LOW	tests/benchdnn/sdpa/ref_sdpa.cpp	288	// Step 4: Softmax over K dimension (axis = 2 of the 3-D score tensor).	COMMENT
LOW	tests/benchdnn/sdpa/ref_sdpa.cpp	306	// Step 5: output = prob_dp x V (matmul primitive).	COMMENT
LOW	third_party/spdlog/spdlog/fmt/bundled/format-inl.h	1260	// Step 1: integer promotion & Schubfach multiplier calculation.	COMMENT
LOW	third_party/spdlog/spdlog/fmt/bundled/format-inl.h	1315	// Step 2: Try larger divisor; remove trailing zeros if necessary.	COMMENT
LOW	third_party/spdlog/spdlog/fmt/bundled/format-inl.h	1347	// Step 3: Find the significand with the smaller divisor.	COMMENT
LOW⚡	examples/matmul_f8_quantization.cpp	257	// Step 1: Compute scaling factors for quantization	COMMENT
LOW⚡	examples/matmul_f8_quantization.cpp	266	// Step 2: Quantize f32 to f8 format with scaling	COMMENT
LOW	examples/matmul_f8_quantization.cpp	340	// Step 3: Matrix multiplication with f8	COMMENT
LOW	examples/matmul_f8_quantization.cpp	383	// Step 4: Validate results	COMMENT
LOW⚡	doc/graph_api/scratchpad.md	33	// Step 1: Query scratchpad requirements	COMMENT
LOW⚡	doc/graph_api/scratchpad.md	36	// Step 2: Check if scratchpad is needed	COMMENT
LOW⚡	doc/graph_api/scratchpad.md	39	// Step 3: Allocate scratchpad tensor	COMMENT
LOW⚡	doc/graph_api/scratchpad.md	44	// Step 4: Execute with user-managed scratchpad	COMMENT
LOW	src/gpu/intel/reduction/utils.cpp	191	// Step 7: Potentially add dst-zero-padding if needed for the final reduction dimensions.	COMMENT
LOW	src/cpu/rv64/rvv_winograd_convolution.cpp	54	// Step 1: temp = G * filter (4x3)	COMMENT
LOW	src/cpu/rv64/rvv_winograd_convolution.cpp	66	// Step 2: result = temp * G^T, store directly to GEMM layout	COMMENT
LOW	src/cpu/rv64/rvv_winograd_convolution.cpp	655	// Step 1: Input transform via JIT kernel (tile loop is inside JIT)	COMMENT
LOW	src/cpu/rv64/rvv_winograd_convolution.cpp	666	// Step 2: GEMM using brgemm kernel (16 elements)	COMMENT
LOW	src/cpu/rv64/rvv_winograd_convolution.cpp	677	// Step 3: Output transform via JIT kernel (tile loop is inside JIT)	COMMENT
LOW⚡	src/cpu/x64/ir/reg_alloc.cpp	593	// Step 1: compute operation-level liveness.	COMMENT
LOW⚡	src/cpu/x64/ir/reg_alloc.cpp	599	// Step 2: build a single live interval per virtual register.	COMMENT
LOW⚡	src/cpu/x64/ir/reg_alloc.cpp	638	// Step 3: weight each virtual register by how often it is referenced and	COMMENT
LOW⚡	src/cpu/x64/ir/reg_alloc.cpp	643	// Step 4: run linear-scan register allocation per physical register file,	COMMENT
LOW⚡	src/common/primitive_desc_iface.cpp	108	// Step 1: create impl::primitive_t or get it from primitive cache	COMMENT
LOW⚡	src/common/primitive_desc_iface.cpp	116	// Step 2: create primitive_iface_t, init and return it to user	COMMENT
LOW⚡	src/common/reorder_primitive_desc_iface.hpp	68	// Step 1: create impl::primitive_t or get it from primitive cache	COMMENT
LOW⚡	src/common/reorder_primitive_desc_iface.hpp	76	// Step 2: create primitive_iface_t, init and return it to user	COMMENT

TODO Padding21 hits · 36 pts

Severity	File	Line	Snippet	Context
LOW	src/gpu/intel/bnorm/nhwc.cpp	240	// TODO: implement it, possible perf boost could be ~ 2x	COMMENT
LOW	src/gpu/intel/bnorm/nhwc_reusable.cpp	149	// TODO: implement it, possible perf boost could be ~ 2x	COMMENT
LOW	src/gpu/intel/bnorm/xe.cpp	180	// TODO: implement it, possible perf boost could be ~ 2x	COMMENT
LOW	src/gpu/intel/gemm/jit/dsl/dsl.cpp	1016	// TODO: Add validation for compatibility between dst/src0/src1.	COMMENT
LOW	src/gpu/intel/rnn/grid.cpp	694	// TODO: implement something like check layout consistency	COMMENT
LOW	src/gpu/intel/rnn/grid.cpp	1650	// TODO: implement without copies	COMMENT
LOW	src/cpu/x64/jit_avx512_core_amx_convolution.cpp	196	// TODO: implement 2D parallelization driver (g * spatial x oc) to increase	COMMENT
LOW	src/cpu/x64/jit_avx512_core_amx_convolution.cpp	582	// TODO: implement 2D parallelization driver (g * spatial x oc) to increase	COMMENT
LOW	src/cpu/x64/jit_avx512_common_conv_kernel.cpp	1171	// TODO: implement a task-based alternative without throttling.	COMMENT
LOW	src/cpu/x64/gemm_bf16_convolution.cpp	220	// TODO: implement store by zmm registers for bf16	COMMENT
LOW	src/cpu/x64/jit_avx512_common_convolution.cpp	99	// TODO: implement it for BWD_D and BWD_W too	COMMENT
LOW	src/cpu/aarch64/jit_sve_conv_kernel.cpp	1105	// TODO: implement a task-based alternative without throttling.	COMMENT
LOW	src/cpu/aarch64/jit_sve_convolution.cpp	101	// TODO: implement it for BWD_D and BWD_W too	COMMENT
LOW	src/cpu/aarch64/brgemm/brgemm.cpp	463	// TODO: implement functionality here similar to corresponding one in x64	COMMENT
LOW	src/cpu/aarch64/matmul/kai_matmul.cpp	221	// TODO: implement zero_points	COMMENT
LOW⚡	src/common/float8.cpp	413	// TODO: implement and use jit conversion kernel for DNNL_X64	COMMENT
LOW⚡	src/common/float8.cpp	422	// TODO: implement and use jit conversion kernel for DNNL_X64	COMMENT
LOW⚡	src/common/float8.cpp	431	// TODO: implement and use jit conversion kernel for DNNL_X64	COMMENT
LOW⚡	src/common/float8.cpp	440	// TODO: implement and use jit conversion kernel for DNNL_X64	COMMENT
LOW⚡	src/common/float8.cpp	450	// TODO: implement and use jit conversion kernel for DNNL_X64	COMMENT
LOW⚡	src/common/float8.cpp	460	// TODO: implement and use jit conversion kernel for DNNL_X64	COMMENT

Deep Nesting17 hits · 17 pts

Severity	File	Line	Context
LOW	scripts/fix_header_guards.py	191	CODE
LOW	scripts/fix_header_guards.py	320	CODE
LOW	scripts/verbose_converter/verbose_converter.py	51	CODE
LOW	scripts/verbose_converter/verbose_converter.py	98	CODE
LOW	scripts/verbose_converter/tests/benchdnn_test.py	51	CODE
LOW	scripts/verbose_converter/src/benchdnn_generator.py	156	CODE
LOW	scripts/verbose_converter/src/benchdnn_generator.py	845	CODE
LOW	scripts/verbose_converter/src/benchdnn_generator.py	881	CODE
LOW	scripts/verbose_converter/src/ir.py	44	CODE
LOW	scripts/verbose_converter/src/parse.py	220	CODE
LOW	scripts/verbose_converter/src/parse.py	263	CODE
LOW	scripts/verbose_converter/src/parse.py	297	CODE
LOW	scripts/verbose_converter/src/parse.py	451	CODE
LOW	scripts/verbose_converter/src/parse.py	703	CODE
LOW	scripts/synthdnn/matmul/primitive.py	261	CODE
LOW	scripts/synthdnn/matmul/primitive.py	281	CODE
LOW	.github/automation/commit-msg-check.py	56	CODE

Empty Error Handlers3 hits · 15 pts

Severity	File	Line	Snippet	Context
HIGH	scripts/verbose_converter/tests/benchdnn_test.py	83	except ValueError:	CODE
HIGH	scripts/verbose_converter/src/parse.py	693	except ValueError:	CODE
HIGH	scripts/verbose_converter/src/parse.py	686	except ValueError:	CODE

Hyper-Verbose Identifiers14 hits · 14 pts

Severity	File	Line	Snippet	Context
LOW	scripts/verbose_converter/src/parse.py	511	def parse_precomputed_reductions(self, prs: str):	CODE
LOW	scripts/verbose_converter/src/parse.py	516	def parse_precomputed_reduction(self, spec) -> ir.PrecomputedReduction:	CODE
LOW	doc/doxyrest/frame/common/doc.lua	511	function getDocBlockListContentsImpl(blockList, context)	CODE
LOW	doc/doxyrest/frame/common/doc.lua	566	function getSimpleDocBlockListContents(blockList)	CODE
LOW	doc/doxyrest/frame/common/string.lua	101	function replaceRolesWithPlainText(source)	CODE
LOW	doc/doxyrest/frame/common/item.lua	124	function getItemInternalDocumentation(item)	CODE
LOW	doc/doxyrest/frame/common/item.lua	137	function getItemBriefDocumentation(item)	CODE
LOW	doc/doxyrest/frame/common/item.lua	160	function getItemDetailedDocumentation(item)	CODE
LOW	doc/doxyrest/frame/common/item.lua	194	function prepareItemArrayDocumentation(itemArray, compound)	CODE
LOW	doc/doxyrest/frame/cfamily/utils.lua	371	function getFunctionModifierDelimiter(indent)	CODE
LOW	doc/doxyrest/frame/cfamily/utils.lua	396	function getFunctionDeclStringImpl(item, returnType, nameTemplate, indent)	CODE
LOW	doc/doxyrest/frame/cfamily/utils.lua	461	function getVoidFunctionDeclString(func, nameTemplate, indent)	CODE
LOW	doc/doxyrest/frame/cfamily/utils.lua	678	function isItemExcludedByProtectionFilter(item)	CODE
LOW	doc/doxyrest/frame/cfamily/utils.lua	698	function isItemExcludedByLocationFilter(item)	CODE

Example Usage Blocks5 hits · 8 pts

Severity	File	Line	Snippet	Context
LOW	include/oneapi/dnnl/dnnl.hpp	4078	/// Example usage:	COMMENT
LOW	include/oneapi/dnnl/dnnl.hpp	4559	/// Example usage:	COMMENT
LOW	include/oneapi/dnnl/dnnl.h	2882	/// Example usage:	COMMENT
LOW	third_party/spdlog/spdlog/mdc.h	19	// Usage example:	COMMENT
LOW	src/graph/utils/pm/pbuilder.hpp	249	// Example usage:	COMMENT

Synthetic Comment Markers1 hit · 8 pts

Severity	File	Line	Snippet	Context
HIGH	tests/benchdnn/utils/cold_cache.cpp	306	// as requested.	COMMENT

AI Response Leakage1 hit · 8 pts

Severity	File	Line	Snippet	Context
HIGH	examples/performance_profiling.cpp	335	/// In this implementation, there is an additional reorder operation that	COMMENT

Slop Phrases2 hits · 6 pts

Severity	File	Line	Snippet	Context
MEDIUM	examples/matmul_f8_quantization.cpp	60	/// @note This is a simplified decoder for demonstration purposes	COMMENT
MEDIUM	examples/matmul_f8_quantization.cpp	90	/// @note This is a simplified decoder for demonstration purposes	COMMENT

Excessive Try-Catch Wrapping2 hits · 2 pts

Severity	File	Line	Snippet	Context
LOW	scripts/verbose_converter/tests/benchdnn_test.py	156	except Exception as e:	CODE
LOW	scripts/verbose_converter/tests/benchdnn_test.py	331	except Exception as e:	STRING

Redundant / Tautological Comments1 hit · 2 pts

Severity	File	Line	Snippet	Context
LOW	scripts/verbose_converter/src/benchdnn_generator.py	194	# Set policy to "host_scalar" if is_host_scalar is True	COMMENT

Unused Imports1 hit · 1 pts

Severity	File	Line	Snippet	Context
LOW	scripts/generate_dnnl_debug.py	18		CODE

Modern Structural Boilerplate1 hit · 1 pts

Severity	File	Line	Snippet	Context
LOW	scripts/verbose_converter/src/parse.py	32	__all__ = ["Parser"]	CODE

uxlfoundation/oneDNN

Open-Source Code Quality, Security and Technical Debt Review