diff --git a/.clang-tidy b/.clang-tidy
index 5776dabe0072..45c5eeb08e8a 100644
--- a/.clang-tidy
+++ b/.clang-tidy
@@ -55,10 +55,12 @@ readability-container-size-empty,
 readability-delete-null-pointer,
 readability-duplicate-include
 readability-misplaced-array-index,
-readability-redundant-function-ptr-dereference,
-readability-redundant-smartptr-get,
+readability-redundant*
 readability-simplify-subscript-expr,
 readability-string-compare,
+-readability-redundant-access-specifiers,
+-readability-redundant-control-flow,
+-readability-redundant-declaration,
 '
 HeaderFilterRegex: '^(aten/|c10/|torch/).*$'
 WarningsAsErrors: '*'
diff --git a/aten/src/ATen/StorageUtils.cpp b/aten/src/ATen/StorageUtils.cpp
index 19c240ed8904..bbd70a4571a3 100644
--- a/aten/src/ATen/StorageUtils.cpp
+++ b/aten/src/ATen/StorageUtils.cpp
@@ -11,7 +11,7 @@ C10_EXPORT c10::intrusive_ptr<c10::StorageImpl> new_shm_fd_storage(
       ALLOCATOR_MAPPED_KEEPFD | ALLOCATOR_MAPPED_UNLINK;
   std::string handle = NewProcessWideShmHandle();
   auto sptr = MapAllocator::makeDataPtr(
-      handle.c_str(), flags, size * sizeof(uint8_t), nullptr);
+      handle, flags, size * sizeof(uint8_t), nullptr);
   return c10::make_intrusive<StorageImpl>(
       c10::StorageImpl::use_byte_size_t(),
       size,
diff --git a/aten/src/ATen/core/Vitals.cpp b/aten/src/ATen/core/Vitals.cpp
index a854be6756bf..13b8eda63859 100644
--- a/aten/src/ATen/core/Vitals.cpp
+++ b/aten/src/ATen/core/Vitals.cpp
@@ -87,7 +87,7 @@ bool APIVitals::setVital(
   return true;
 }
 
-APIVitals::APIVitals() : vitals_enabled(false), name_map_() {
+APIVitals::APIVitals() : vitals_enabled(false) {
   // Set default values, force is necessary because in unit tests the env
   // variable may not be set when global APIVitals are constructed.
   setVital("CUDA", "used", "False", /* force = */ true);
diff --git a/aten/src/ATen/core/Vitals.h b/aten/src/ATen/core/Vitals.h
index 7ec213938d56..2fd7729744a1 100644
--- a/aten/src/ATen/core/Vitals.h
+++ b/aten/src/ATen/core/Vitals.h
@@ -11,7 +11,7 @@ TORCH_API bool torchVitalEnabled();
 
 struct TORCH_API TorchVitalAttr {
   // always initialized to empty
-  std::string value = "";
+  std::string value;
   template <typename T>
   TorchVitalAttr& operator<<(const T& t) {
     if (torchVitalEnabled()) {
diff --git a/aten/src/ATen/core/blob.h b/aten/src/ATen/core/blob.h
index c45679794045..251da65e0896 100644
--- a/aten/src/ATen/core/blob.h
+++ b/aten/src/ATen/core/blob.h
@@ -22,7 +22,7 @@ class TORCH_API Blob final : public c10::intrusive_ptr_target {
   /**
    * Initializes an empty Blob.
    */
-  Blob() noexcept : meta_() {}
+  Blob() noexcept = default;
   ~Blob() override {
     Reset();
   }
diff --git a/aten/src/ATen/core/class_type.cpp b/aten/src/ATen/core/class_type.cpp
index 7fb0d355529a..800d9ea0ef9f 100644
--- a/aten/src/ATen/core/class_type.cpp
+++ b/aten/src/ATen/core/class_type.cpp
@@ -76,7 +76,7 @@ std::string ClassType::getForwardPreHookErrorMessage(size_t pre_hook_idx) const
   std::string input_types = getSchemaInputTypesString(forward_schema);
   const std::vector<Argument>& forward_args = forward_schema.arguments();
 
-  std::string single_output = "";
+  std::string single_output;
   if (forward_args.size() == 2 &&
       forward_args[1].type()->cast<TupleType>() == nullptr) {
     // if the output type is a single tuple, it needs to be wrapped in an outer tuple
diff --git a/aten/src/ATen/core/class_type.h b/aten/src/ATen/core/class_type.h
index d3373fd2ee38..ea124fc6eb07 100644
--- a/aten/src/ATen/core/class_type.h
+++ b/aten/src/ATen/core/class_type.h
@@ -432,7 +432,7 @@ struct TORCH_API ClassType : public NamedType {
   bool isModule_ = false;
 
   // Doc string of class.
-  std::string doc_string_ = "";
+  std::string doc_string_;
 
   // For error reporting accesses to class level attributes.
   std::vector<std::string> unresolved_class_attributes_;
diff --git a/aten/src/ATen/core/function.h b/aten/src/ATen/core/function.h
index cebc10640a4c..7e8a765a05ab 100644
--- a/aten/src/ATen/core/function.h
+++ b/aten/src/ATen/core/function.h
@@ -43,7 +43,7 @@ struct TORCH_API Function {
   Function(Function&&) noexcept = default;
   Function& operator=(Function&&) noexcept = default;
   virtual std::string_view doc_string() const {
-    static constexpr std::string_view no_doc_string = "";
+    static constexpr std::string_view no_doc_string;
     return no_doc_string;
   }
 
diff --git a/aten/src/ATen/core/function_schema.h b/aten/src/ATen/core/function_schema.h
index 3ec771328110..48685c62d6cd 100644
--- a/aten/src/ATen/core/function_schema.h
+++ b/aten/src/ATen/core/function_schema.h
@@ -567,7 +567,7 @@ inline std::ostream& operator<<(std::ostream& out, const Argument& arg) {
     if (arg.alias_info() && !arg.alias_info()->containedTypes().empty()){
       out << arg.alias_info()->containedTypes()[0];
     }
-    std::string N = "";
+    std::string N;
     if (arg.N()) {
         N = std::to_string(*arg.N());
     }
diff --git a/aten/src/ATen/core/jit_type.h b/aten/src/ATen/core/jit_type.h
index c5cbe512fc2a..2c8fbf8a09ad 100644
--- a/aten/src/ATen/core/jit_type.h
+++ b/aten/src/ATen/core/jit_type.h
@@ -2245,7 +2245,7 @@ static const TypeKind Kind = TypeKind::ScalarTypeType;
 static ScalarTypeTypePtr get();
 
 private:
-ScalarTypeType() : EnumerationType() {}
+ScalarTypeType()  {}
 };
 
 struct MemoryFormatType;
@@ -2259,7 +2259,7 @@ static const TypeKind Kind = TypeKind::MemoryFormatType;
 static MemoryFormatTypePtr get();
 
 private:
-MemoryFormatType() : EnumerationType() {}
+MemoryFormatType()  {}
 };
 
 struct LayoutType;
@@ -2273,7 +2273,7 @@ static const TypeKind Kind = TypeKind::LayoutType;
 static LayoutTypePtr get();
 
 private:
-LayoutType() : EnumerationType() {}
+LayoutType()  {}
 };
 
 namespace detail {
diff --git a/aten/src/ATen/core/library.cpp b/aten/src/ATen/core/library.cpp
index 8657cd9274f8..b8a5b418bbc0 100644
--- a/aten/src/ATen/core/library.cpp
+++ b/aten/src/ATen/core/library.cpp
@@ -48,7 +48,6 @@ CppFunction::CppFunction(c10::KernelFunction func, std::optional<c10::impl::CppS
   : func_(std::move(func))
   , cpp_signature_(cpp_signature)
   , schema_(std::move(schema))
-  , debug_()
   {}
 
 CppFunction::~CppFunction() = default;
diff --git a/c10/core/Allocator.h b/c10/core/Allocator.h
index c881d104934b..fed5fbc6e220 100644
--- a/c10/core/Allocator.h
+++ b/c10/core/Allocator.h
@@ -31,7 +31,7 @@ class C10_API DataPtr {
  public:
   // Choice of CPU here is arbitrary; if there's an "undefined" device
   // we could use that too
-  DataPtr() : ptr_(), device_(DeviceType::CPU) {}
+  DataPtr() : device_(DeviceType::CPU) {}
   DataPtr(void* data, Device device) : ptr_(data), device_(device) {}
   DataPtr(void* data, void* ctx, DeleterFnPtr ctx_deleter, Device device)
       : ptr_(data, ctx, ctx_deleter), device_(device) {}
diff --git a/c10/cuda/CUDAAllocatorConfig.cpp b/c10/cuda/CUDAAllocatorConfig.cpp
index 7c1c1e02644b..0f27cd08d070 100644
--- a/c10/cuda/CUDAAllocatorConfig.cpp
+++ b/c10/cuda/CUDAAllocatorConfig.cpp
@@ -18,8 +18,7 @@ CUDAAllocatorConfig::CUDAAllocatorConfig()
       m_expandable_segments(false),
       m_release_lock_on_cudamalloc(false),
       m_pinned_use_cuda_host_register(false),
-      m_pinned_use_background_threads(false),
-      m_last_allocator_settings("") {
+      m_pinned_use_background_threads(false) {
   m_roundup_power2_divisions.assign(kRoundUpPowerOfTwoIntervals, 0);
 }
 
diff --git a/c10/cuda/CUDACachingAllocator.cpp b/c10/cuda/CUDACachingAllocator.cpp
index 03174cf9ad03..93e915939f76 100644
--- a/c10/cuda/CUDACachingAllocator.cpp
+++ b/c10/cuda/CUDACachingAllocator.cpp
@@ -211,7 +211,6 @@ struct Block {
       void* ptr)
       : device(device),
         stream(stream),
-        stream_uses(),
         size(size),
         requested_size(0),
         pool(pool),
@@ -219,11 +218,7 @@ struct Block {
 
   // constructor for search key
   Block(c10::DeviceIndex device, cudaStream_t stream, size_t size)
-      : device(device),
-        stream(stream),
-        stream_uses(),
-        size(size),
-        requested_size(0) {}
+      : device(device), stream(stream), size(size), requested_size(0) {}
 
   size_t gc_count() {
     TORCH_INTERNAL_ASSERT(pool);
diff --git a/c10/cuda/CUDAGuard.h b/c10/cuda/CUDAGuard.h
index 0a076d5637b1..0fb0e737a4d3 100644
--- a/c10/cuda/CUDAGuard.h
+++ b/c10/cuda/CUDAGuard.h
@@ -74,7 +74,7 @@ struct CUDAGuard {
 /// CUDAGuard for when you can use this.
 struct OptionalCUDAGuard {
   /// Create an uninitialized OptionalCUDAGuard.
-  explicit OptionalCUDAGuard() : guard_() {}
+  explicit OptionalCUDAGuard() = default;
 
   /// Set the current CUDA device to the passed Device, if it is not nullopt.
   explicit OptionalCUDAGuard(std::optional<Device> device_opt)
@@ -208,7 +208,7 @@ struct CUDAStreamGuard {
 /// CUDAGuard for when you can use this.
 struct OptionalCUDAStreamGuard {
   /// Create an uninitialized guard.
-  explicit OptionalCUDAStreamGuard() : guard_() {}
+  explicit OptionalCUDAStreamGuard() = default;
 
   /// Set the current CUDA device to the device associated with the passed
   /// stream, and set the current CUDA stream on that device to the passed
diff --git a/c10/test/util/string_util_test.cpp b/c10/test/util/string_util_test.cpp
index 963253a4fcb0..62c053028f70 100644
--- a/c10/test/util/string_util_test.cpp
+++ b/c10/test/util/string_util_test.cpp
@@ -49,8 +49,8 @@ TEST(StringUtilTest, testStrWideSingleMultibyte) {
 
 namespace test_str_wide_empty {
 TEST(StringUtilTest, testStrWideEmpty) {
-  std::wstring s = L"";
-  std::string narrow = "";
+  std::wstring s;
+  std::string narrow;
   EXPECT_EQ(narrow, c10::str(s));
 
   const wchar_t* c_str = s.c_str();
diff --git a/c10/util/LeftRight.h b/c10/util/LeftRight.h
index 0ad9a1b34610..759f6967933d 100644
--- a/c10/util/LeftRight.h
+++ b/c10/util/LeftRight.h
@@ -57,8 +57,7 @@ class LeftRight final {
       : _counters{{{0}, {0}}},
         _foregroundCounterIndex(0),
         _foregroundDataIndex(0),
-        _data{{T{args...}, T{args...}}},
-        _writeMutex() {}
+        _data{{T{args...}, T{args...}}} {}
 
   // Copying and moving would not be threadsafe.
   // Needs more thought and careful design to make that work.
diff --git a/c10/util/signal_handler.cpp b/c10/util/signal_handler.cpp
index 7132f08588ce..267fc5721b28 100644
--- a/c10/util/signal_handler.cpp
+++ b/c10/util/signal_handler.cpp
@@ -112,8 +112,6 @@ FatalSignalHandler::FatalSignalHandler()
     : fatalSignalHandlersInstalled(false),
       fatalSignalReceived(false),
       fatalSignalName("<UNKNOWN>"),
-      writingCond(),
-      writingMutex(),
       signalReceived(false) {}
 
 // NOLINTNEXTLINE(cppcoreguidelines-avoid-c-arrays,modernize-avoid-c-arrays)
diff --git a/torch/csrc/CudaIPCTypes.cpp b/torch/csrc/CudaIPCTypes.cpp
index cf1c7217045e..88875516f43f 100644
--- a/torch/csrc/CudaIPCTypes.cpp
+++ b/torch/csrc/CudaIPCTypes.cpp
@@ -150,7 +150,6 @@ CudaIPCSentData::CudaIPCSentData(
     : handle_(std::move(handle)),
       offset_(offset),
       counter_ptr_(counter_ptr),
-      original_ptr_(),
       device_(device) {
 #if !defined(USE_ROCM)
   // CUDA have the unofficial limit on the number of recorded blocking
diff --git a/torch/csrc/api/include/torch/detail/TensorDataContainer.h b/torch/csrc/api/include/torch/detail/TensorDataContainer.h
index d5e8f0f9234b..5544bbc0a0b9 100644
--- a/torch/csrc/api/include/torch/detail/TensorDataContainer.h
+++ b/torch/csrc/api/include/torch/detail/TensorDataContainer.h
@@ -126,8 +126,7 @@ struct TensorDataContainer {
   AT_FORALL_COMPLEX_TYPES(TENSOR)
 #undef TENSOR
   TensorDataContainer(std::initializer_list<TensorDataContainer> init_list)
-      : sizes_(),
-        scalar_type_(init_list.begin()->scalar_type()),
+      : scalar_type_(init_list.begin()->scalar_type()),
         type_(TensorDataContainerType::InitList),
         init_list_(init_list) {
     const TensorDataContainer& first_elem = *(init_list.begin());
diff --git a/torch/csrc/api/include/torch/nn/modules/container/sequential.h b/torch/csrc/api/include/torch/nn/modules/container/sequential.h
index f5ddb4e370f6..e45a6af29401 100644
--- a/torch/csrc/api/include/torch/nn/modules/container/sequential.h
+++ b/torch/csrc/api/include/torch/nn/modules/container/sequential.h
@@ -376,7 +376,7 @@ class Sequential : public torch::nn::ModuleHolder<SequentialImpl> {
  public:
   using torch::nn::ModuleHolder<SequentialImpl>::ModuleHolder;
 
-  Sequential() : ModuleHolder() {}
+  Sequential() {}
 
   /// Constructs the `Sequential` from a braced-init-list of named `AnyModule`s.
   /// It enables the following use case:
diff --git a/torch/csrc/autograd/function.h b/torch/csrc/autograd/function.h
index 6e711b384cb5..b5360c6764f2 100644
--- a/torch/csrc/autograd/function.h
+++ b/torch/csrc/autograd/function.h
@@ -766,7 +766,7 @@ edge_list collect_next_edges(Variables&&... variables) {
 }
 
 struct TypeAndSize {
-  TypeAndSize() : options(at::TensorOptions()) {}
+  TypeAndSize() {}
   /* implicit */
   TypeAndSize(const at::Tensor& t)
       : sym_sizes(t.sym_sizes().vec()), options(t.options()) {}
diff --git a/torch/csrc/autograd/functions/tensor.cpp b/torch/csrc/autograd/functions/tensor.cpp
index cd271db75bf6..1c13b60ca783 100644
--- a/torch/csrc/autograd/functions/tensor.cpp
+++ b/torch/csrc/autograd/functions/tensor.cpp
@@ -65,8 +65,7 @@ CopySlices::CopySlices(
     at::TensorGeometry view_,
     std::unique_ptr<ViewFunc> view_fn_,
     std::shared_ptr<Node> fn_)
-    : Node(),
-      base(base_var),
+    : base(base_var),
       view(std::move(view_)),
       view_fn(std::move(view_fn_)),
       fn(std::move(fn_)) {
diff --git a/torch/csrc/autograd/profiler_python.cpp b/torch/csrc/autograd/profiler_python.cpp
index f1d2822aeac6..51d1cd355563 100644
--- a/torch/csrc/autograd/profiler_python.cpp
+++ b/torch/csrc/autograd/profiler_python.cpp
@@ -603,8 +603,7 @@ static PyTypeObject TraceContextType = {
 
 class gil_and_restore_thread {
  public:
-  gil_and_restore_thread()
-      : gil_(), initial_thread_state_{PyThreadState_Get()} {}
+  gil_and_restore_thread() : initial_thread_state_{PyThreadState_Get()} {}
   ~gil_and_restore_thread() {
     PyThreadState_Swap(initial_thread_state_);
 
diff --git a/torch/csrc/distributed/autograd/engine/dist_engine.cpp b/torch/csrc/distributed/autograd/engine/dist_engine.cpp
index 4f8dfd6456df..6c3b3537c523 100644
--- a/torch/csrc/distributed/autograd/engine/dist_engine.cpp
+++ b/torch/csrc/distributed/autograd/engine/dist_engine.cpp
@@ -108,8 +108,7 @@ void DistEngine::globalCpuThread(
 }
 
 DistEngine::DistEngine()
-    : initializedContextIds_(),
-      engine_(Engine::get_default_engine()),
+    : engine_(Engine::get_default_engine()),
       global_cpu_ready_queue_(std::make_shared<ReadyQueue>()),
       global_cpu_thread_(
           &DistEngine::globalCpuThread,
diff --git a/torch/csrc/distributed/c10d/FileStore.cpp b/torch/csrc/distributed/c10d/FileStore.cpp
index 3b4c38711f55..efd8fc647bb8 100644
--- a/torch/csrc/distributed/c10d/FileStore.cpp
+++ b/torch/csrc/distributed/c10d/FileStore.cpp
@@ -282,8 +282,7 @@ off_t refresh(
 } // namespace
 
 FileStore::FileStore(std::string path, int numWorkers)
-    : Store(),
-      path_(std::move(path)),
+    : path_(std::move(path)),
 
       numWorkers_(numWorkers),
       cleanupKey_("cleanup/"),
diff --git a/torch/csrc/distributed/c10d/Work.cpp b/torch/csrc/distributed/c10d/Work.cpp
index 4502e4aa235b..5e30e91ce05b 100644
--- a/torch/csrc/distributed/c10d/Work.cpp
+++ b/torch/csrc/distributed/c10d/Work.cpp
@@ -147,7 +147,7 @@ uint64_t Work::getSequencenumber() const {
 class FutureWrappingWork : public Work {
  public:
   FutureWrappingWork(c10::intrusive_ptr<c10::ivalue::Future> fut)
-      : Work(), _fut(std::move(fut)) {}
+      : _fut(std::move(fut)) {}
 
   ~FutureWrappingWork() override = default;
 
diff --git a/torch/csrc/distributed/rpc/rref_impl.cpp b/torch/csrc/distributed/rpc/rref_impl.cpp
index 937d491a18b2..a004a5e2b1c1 100644
--- a/torch/csrc/distributed/rpc/rref_impl.cpp
+++ b/torch/csrc/distributed/rpc/rref_impl.cpp
@@ -53,10 +53,7 @@ RRefForkData::RRefForkData(
 //////////////////////////////  RRef  /////////////////////////////////////
 
 RRef::RRef(worker_id_t ownerId, const RRefId& rrefId, TypePtr type)
-    : RRefInterface(),
-      ownerId_(ownerId),
-      rrefId_(rrefId),
-      type_(std::move(type)) {}
+    : ownerId_(ownerId), rrefId_(rrefId), type_(std::move(type)) {}
 
 RRefForkData RRef::fork() const {
   auto& ctx = RRefContext::getInstance();
diff --git a/torch/csrc/dynamo/guards.cpp b/torch/csrc/dynamo/guards.cpp
index d6458d506568..b21af7beb760 100644
--- a/torch/csrc/dynamo/guards.cpp
+++ b/torch/csrc/dynamo/guards.cpp
@@ -3109,7 +3109,7 @@ class TORCH_FUNCTION_MODE_STACK : public LeafGuard {
   TORCH_FUNCTION_MODE_STACK(
       const py::list& initial_stack,
       py::object verbose_code_parts)
-      : LeafGuard(std::move(verbose_code_parts)), _ref_stack() {
+      : LeafGuard(std::move(verbose_code_parts)) {
     Py_ssize_t len = PyList_Size(initial_stack.ptr());
     for (Py_ssize_t idx = 0; idx < len; idx++) {
       PyObject* mode = PyList_GetItem(initial_stack.ptr(), idx); // borrowed ref
diff --git a/torch/csrc/inductor/aoti_eager/kernel_holder.h b/torch/csrc/inductor/aoti_eager/kernel_holder.h
index fed2e3b5d61d..8459b35c6837 100644
--- a/torch/csrc/inductor/aoti_eager/kernel_holder.h
+++ b/torch/csrc/inductor/aoti_eager/kernel_holder.h
@@ -21,7 +21,7 @@ struct AOTIKernelMetadata {
   std::vector<ParameterMetadata> parameter_metadata_list_;
   // AOTI model runner to run the AOTI kernel
   std::shared_ptr<AOTIModelContainerRunner> kernel_runner_;
-  AOTIKernelMetadata() : parameter_metadata_list_(), kernel_runner_(nullptr) {}
+  AOTIKernelMetadata() : kernel_runner_(nullptr) {}
 
   // Check whether the given parameter metadata list is the same as the
   // parameter metadata list of the AOTI kernel.
diff --git a/torch/csrc/inductor/aoti_package/model_package_loader.cpp b/torch/csrc/inductor/aoti_package/model_package_loader.cpp
index 9473492bb305..3dbb8b810aa4 100644
--- a/torch/csrc/inductor/aoti_package/model_package_loader.cpp
+++ b/torch/csrc/inductor/aoti_package/model_package_loader.cpp
@@ -91,7 +91,7 @@ std::tuple<std::string, std::string> get_cpp_compile_command(
   std::string compiler = compile_options["compiler"].get<std::string>();
   bool compile_only = compile_options["compile_only"].get<bool>();
 
-  std::string source_args = "";
+  std::string source_args;
   for (const std::string& source : sources) {
     source_args += source + " ";
   }
@@ -99,37 +99,37 @@ std::tuple<std::string, std::string> get_cpp_compile_command(
   std::string file_ext = compile_only ? ".o" : ".so";
   std::string target_file = output_dir + filename + file_ext;
 
-  std::string cflags_args = "";
+  std::string cflags_args;
   for (auto& arg : compile_options["cflags"]) {
     cflags_args += "-" + arg.get<std::string>() + " ";
   }
 
-  std::string definitions_args = "";
+  std::string definitions_args;
   for (auto& arg : compile_options["definitions"]) {
     definitions_args += "-D " + arg.get<std::string>() + " ";
   }
 
-  std::string include_dirs_args = "";
+  std::string include_dirs_args;
   for (auto& arg : compile_options["include_dirs"]) {
     include_dirs_args += "-I" + arg.get<std::string>() + " ";
   }
 
-  std::string ldflags_args = "";
+  std::string ldflags_args;
   for (auto& arg : compile_options["ldflags"]) {
     ldflags_args += "-" + arg.get<std::string>() + " ";
   }
 
-  std::string libraries_dirs_args = "";
+  std::string libraries_dirs_args;
   for (auto& arg : compile_options["libraries_dirs"]) {
     libraries_dirs_args += "-L" + arg.get<std::string>() + " ";
   }
 
-  std::string libraries_args = "";
+  std::string libraries_args;
   for (auto& arg : compile_options["libraries"]) {
     libraries_args += "-l" + arg.get<std::string>() + " ";
   }
 
-  std::string passthrough_parameters_args = "";
+  std::string passthrough_parameters_args;
   for (auto& arg : compile_options["passthrough_args"]) {
     passthrough_parameters_args += arg.get<std::string>() + " ";
   }
@@ -343,10 +343,10 @@ AOTIModelPackageLoader::AOTIModelPackageLoader(
   }
 
   temp_dir_ = create_temp_dir();
-  std::string so_filename = "";
-  std::string cpp_filename = "";
-  std::string consts_filename = "";
-  std::string found_filenames = ""; // Saving for bookkeeping
+  std::string so_filename;
+  std::string cpp_filename;
+  std::string consts_filename;
+  std::string found_filenames; // Saving for bookkeeping
   std::string model_directory =
       "data" + k_separator + "aotinductor" + k_separator + model_name;
 
@@ -379,7 +379,7 @@ AOTIModelPackageLoader::AOTIModelPackageLoader(
             "Failed to find parent path in " + output_path_str);
       }
       std::string parent_path = output_path_str.substr(0, parent_path_idx);
-      if (!recursive_mkdir(parent_path.c_str())) {
+      if (!recursive_mkdir(parent_path)) {
         throw std::runtime_error(fmt::format(
             "Failed to create directory {}: {}",
             parent_path,
diff --git a/torch/csrc/lazy/core/hash.h b/torch/csrc/lazy/core/hash.h
index 0ea3022f2696..d9143da8beb8 100644
--- a/torch/csrc/lazy/core/hash.h
+++ b/torch/csrc/lazy/core/hash.h
@@ -29,7 +29,7 @@ class TORCH_API hash_t : public c10::uint128 {
   hash_t(uint64_t val) : uint128(val) {}
   hash_t(uint128 val) : uint128(val) {}
   hash_t(uint64_t top, uint64_t bottom) : uint128(top, bottom) {}
-  hash_t() : uint128() {}
+  hash_t() {}
 };
 
 // Std* functions use 64-bit hash
diff --git a/torch/csrc/profiler/standalone/execution_trace_observer.cpp b/torch/csrc/profiler/standalone/execution_trace_observer.cpp
index 1f6953667c99..1990c30c73bf 100644
--- a/torch/csrc/profiler/standalone/execution_trace_observer.cpp
+++ b/torch/csrc/profiler/standalone/execution_trace_observer.cpp
@@ -410,7 +410,7 @@ convertIValue(
     size_t offset = 0;
     size_t numel = 0;
     size_t itemsize = 0;
-    std::string device_str = "";
+    std::string device_str;
     // symbolic sizes/strides implies t->storage_offset() will fail
     if (tensor_impl->has_storage() &&
         !tensor_impl->has_symbolic_sizes_strides()) {
diff --git a/torch/csrc/utils/invalid_arguments.cpp b/torch/csrc/utils/invalid_arguments.cpp
index c2825f7d945d..a0a85956442d 100644
--- a/torch/csrc/utils/invalid_arguments.cpp
+++ b/torch/csrc/utils/invalid_arguments.cpp
@@ -111,7 +111,7 @@ struct Option {
         is_variadic(is_variadic),
         has_out(has_out) {}
   Option(bool is_variadic, bool has_out)
-      : arguments(), is_variadic(is_variadic), has_out(has_out) {}
+      : is_variadic(is_variadic), has_out(has_out) {}
   Option(const Option&) = delete;
   Option(Option&& other) noexcept = default;
   Option& operator=(const Option&) = delete;
diff --git a/torch/csrc/utils/throughput_benchmark.h b/torch/csrc/utils/throughput_benchmark.h
index 50854f1b73aa..8cf2f97158f2 100644
--- a/torch/csrc/utils/throughput_benchmark.h
+++ b/torch/csrc/utils/throughput_benchmark.h
@@ -58,7 +58,7 @@ struct BenchmarkConfig {
   // If set autograd profiler will be enabled. I.e. this variable would be
   // created before the main benchmark loop (but after the warmup):
   // RecordProfile guard(profiler_output_path);
-  std::string profiler_output_path{""};
+  std::string profiler_output_path;
 };
 
 namespace detail {
diff --git a/torch/library.h b/torch/library.h
index 2761573e2ccc..4914f3d4189e 100644
--- a/torch/library.h
+++ b/torch/library.h
@@ -119,8 +119,8 @@ class TORCH_API CppFunction final {
       : func_(c10::KernelFunction::makeFromUnboxedRuntimeFunction(f)),
         cpp_signature_(c10::impl::CppSignature::make<Func>()),
         schema_(
-            c10::detail::inferFunctionSchemaFromFunctor<std::decay_t<Func>>()),
-        debug_() {}
+            c10::detail::inferFunctionSchemaFromFunctor<std::decay_t<Func>>())
+        {}
 
   /// This overload accepts compile time function pointers, e.g.,
   /// `CppFunction(TORCH_FN(add_impl))`
@@ -134,8 +134,8 @@ class TORCH_API CppFunction final {
         cpp_signature_(
             c10::impl::CppSignature::make<typename FuncPtr::FuncType>()),
         schema_(c10::detail::inferFunctionSchemaFromFunctor<
-                typename FuncPtr::FuncType>()),
-        debug_() {}
+                typename FuncPtr::FuncType>())
+        {}
 
   /// This overload accepts lambdas, e.g., `CppFunction([](const Tensor& self) {
   /// ... })`
@@ -149,8 +149,8 @@ class TORCH_API CppFunction final {
             std::forward<Lambda>(f))),
         cpp_signature_(c10::impl::CppSignature::make<Lambda>()),
         schema_(c10::detail::inferFunctionSchemaFromFunctor<
-                std::decay_t<Lambda>>()),
-        debug_() {}
+                std::decay_t<Lambda>>())
+        {}
 
 #if defined C10_MOBILE
   /// This overload accepts function pointers, e.g., `CppFunction(&add_impl,