diff --git a/backends/arm/scripts/docgen/generate_op_support.py b/backends/arm/scripts/docgen/generate_op_support.py
index 5b1b9c79d5b..c48ecb3166f 100644
--- a/backends/arm/scripts/docgen/generate_op_support.py
+++ b/backends/arm/scripts/docgen/generate_op_support.py
@@ -356,7 +356,6 @@ def _activate_backend(backend: str | BackendConfig) -> BackendConfig:
     "torch.ops.aten.native_layer_norm.default",
     "torch.ops.aten.embedding.default",
     "torch.ops.aten.pow.Tensor_Tensor",
-    "torch.ops.aten.scaled_dot_product_attention.default",
 }
 
 
@@ -405,18 +404,6 @@ def _activate_backend(backend: str | BackendConfig) -> BackendConfig:
     ): {
         "INT": {"torch.ops.aten.embedding.default"},
     },
-    (
-        "backends/arm/test/ops/test_sdpa.py",
-        "test_sdpa_vgf_no_quant",
-    ): {
-        "FP": {"torch.ops.aten.scaled_dot_product_attention.default"},
-    },
-    (
-        "backends/arm/test/ops/test_sdpa.py",
-        "test_sdpa_vgf_quant",
-    ): {
-        "INT": {"torch.ops.aten.scaled_dot_product_attention.default"},
-    },
 }
 
 # Existing U55 runtime tests below intentionally suppress direct ATen/Edge
@@ -459,12 +446,6 @@ def _activate_backend(backend: str | BackendConfig) -> BackendConfig:
     ): {
         "INT": {"torch.ops.aten.pow.Tensor_Tensor"},
     },
-    (
-        "backends/arm/test/ops/test_sdpa.py",
-        "test_sdpa_u55_INT",
-    ): {
-        "INT": {"torch.ops.aten.scaled_dot_product_attention.default"},
-    },
 }
 
 # Existing U85 runtime tests below intentionally suppress direct ATen/Edge
@@ -497,12 +478,6 @@ def _activate_backend(backend: str | BackendConfig) -> BackendConfig:
     ): {
         "INT": {"torch.ops.aten.pow.Tensor_Tensor"},
     },
-    (
-        "backends/arm/test/ops/test_sdpa.py",
-        "test_sdpa_u85_INT",
-    ): {
-        "INT": {"torch.ops.aten.scaled_dot_product_attention.default"},
-    },
 }
 
 
@@ -635,9 +610,6 @@ def _active_explicit_backend_coverage() -> dict[tuple[str, str], dict[str, set[s
         "torch.nn.Embedding",
         "torch.nn.functional.embedding",
     ),
-    "torch.ops.aten.scaled_dot_product_attention.default": (
-        "torch.nn.functional.scaled_dot_product_attention",
-    ),
     # Activations.
     "torch.ops.aten.relu.default": ("torch.relu", "torch.nn.ReLU"),
     "torch.ops.aten.relu_.default": ("torch.Tensor.relu_",),
diff --git a/backends/arm/test/misc/test_docgen_op_support.py b/backends/arm/test/misc/test_docgen_op_support.py
index bf15667a8a5..6c2916dd0ec 100644
--- a/backends/arm/test/misc/test_docgen_op_support.py
+++ b/backends/arm/test/misc/test_docgen_op_support.py
@@ -215,39 +215,6 @@ def test_u85_explicit_coverage_attribution() -> None:
         docgen._activate_backend(original)
 
 
-SDPA_OP = "torch.ops.aten.scaled_dot_product_attention.default"
-
-
-def test_sdpa_is_decomposed_with_public_api_alias() -> None:
-    assert SDPA_OP in docgen.DECOMPOSED_OPS
-    assert docgen._pytorch_api_aliases(SDPA_OP) == (
-        "torch.nn.functional.scaled_dot_product_attention",
-    )
-
-
-@pytest.mark.parametrize(
-    ("backend", "function", "profile"),
-    [
-        ("vgf", "test_sdpa_vgf_no_quant", "FP"),
-        ("vgf", "test_sdpa_vgf_quant", "INT"),
-        ("u55", "test_sdpa_u55_INT", "INT"),
-        ("u85", "test_sdpa_u85_INT", "INT"),
-    ],
-)
-def test_sdpa_explicit_backend_coverage(
-    backend: str, function: str, profile: str
-) -> None:
-    original = docgen.ACTIVE_BACKEND_KEY
-    try:
-        docgen._activate_backend(backend)
-        coverage = docgen._active_explicit_backend_coverage()
-        assert coverage[("backends/arm/test/ops/test_sdpa.py", function)][profile] == {
-            SDPA_OP
-        }
-    finally:
-        docgen._activate_backend(original)
-
-
 def test_non_vgf_backend_does_not_collect_vgf_custom_partition_ops() -> None:
     original = docgen.ACTIVE_BACKEND_KEY
     try:
diff --git a/docs/source/backends/arm-ethos-u/U55_op_support.md b/docs/source/backends/arm-ethos-u/U55_op_support.md
index 5b6a708546e..0a03854d877 100644
--- a/docs/source/backends/arm-ethos-u/U55_op_support.md
+++ b/docs/source/backends/arm-ethos-u/U55_op_support.md
@@ -6,7 +6,7 @@ This page lists Ethos-U55-supported PyTorch APIs and the dtype and quantization
 
 `8x8` means 8-bit activations and 8-bit weights. `16x8` means 16-bit activations and 8-bit weights. `8x4` means 8-bit activations and 4-bit weights.
 
-Total supported PyTorch APIs: **105**.
+Total supported PyTorch APIs: **104**.
 
 | PyTorch API | Support profile | DType | Quantization mode |
 | --- | --- | --- | --- |
@@ -68,7 +68,6 @@ Total supported PyTorch APIs: **105**.
 | `torch.nn.Conv2d` / `torch.nn.functional.conv2d` | INT | `INT8`, `INT16`, `INT4` | 8x8, 8x4, 16x8 |
 | `torch.nn.ELU` / `torch.nn.functional.elu` | INT | `INT8` | 8x8 |
 | `torch.nn.functional.pad` | INT | `INT8` | 8x8 |
-| `torch.nn.functional.scaled_dot_product_attention` | INT | `INT8` | 8x8 |
 | `torch.nn.GELU` / `torch.nn.functional.gelu` | INT | `INT8`, `INT16` | 8x8, 16x8 |
 | `torch.nn.Hardsigmoid` / `torch.nn.functional.hardsigmoid` | INT | `INT8` | 8x8 |
 | `torch.nn.Hardswish` / `torch.nn.functional.hardswish` | INT | `INT8` | 8x8 |
diff --git a/docs/source/backends/arm-ethos-u/U85_op_support.md b/docs/source/backends/arm-ethos-u/U85_op_support.md
index 204d411d4cc..7d778ff5cc9 100644
--- a/docs/source/backends/arm-ethos-u/U85_op_support.md
+++ b/docs/source/backends/arm-ethos-u/U85_op_support.md
@@ -6,7 +6,7 @@ This page lists Ethos-U85-supported PyTorch APIs and the dtype and quantization
 
 `8x8` means 8-bit activations and 8-bit weights. `16x8` means 16-bit activations and 8-bit weights. `8x4` means 8-bit activations and 4-bit weights.
 
-Total supported PyTorch APIs: **128**.
+Total supported PyTorch APIs: **127**.
 
 | PyTorch API | Support profile | DType | Quantization mode |
 | --- | --- | --- | --- |
@@ -87,7 +87,6 @@ Total supported PyTorch APIs: **128**.
 | `torch.nn.ELU` / `torch.nn.functional.elu` | INT | `INT8` | 8x8 |
 | `torch.nn.Embedding` / `torch.nn.functional.embedding` | INT | `INT8` | 8x8 |
 | `torch.nn.functional.pad` | INT | `INT8` | 8x8 |
-| `torch.nn.functional.scaled_dot_product_attention` | INT | `INT8` | 8x8 |
 | `torch.nn.GELU` / `torch.nn.functional.gelu` | INT | `INT8`, `INT16` | 8x8, 16x8 |
 | `torch.nn.Hardsigmoid` / `torch.nn.functional.hardsigmoid` | INT | `INT8` | 8x8 |
 | `torch.nn.Hardswish` / `torch.nn.functional.hardswish` | INT | `INT8` | 8x8 |
diff --git a/docs/source/backends/arm-vgf/VGF_op_support.md b/docs/source/backends/arm-vgf/VGF_op_support.md
index 368eff8c6d1..4a2bc8f8816 100644
--- a/docs/source/backends/arm-vgf/VGF_op_support.md
+++ b/docs/source/backends/arm-vgf/VGF_op_support.md
@@ -105,7 +105,6 @@ Total supported PyTorch APIs: **158**.
 | `torch.nn.ELU` / `torch.nn.functional.elu` | FP, INT | `FP32`, `INT8` | 8x8 |
 | `torch.nn.Embedding` / `torch.nn.functional.embedding` | FP, INT | `FP32`, `INT8` | 8x8 |
 | `torch.nn.functional.pad` | FP, INT | `FP32`, `FP16`, `BF16`, `INT8`, `INT16` | 8x8, 16x8 |
-| `torch.nn.functional.scaled_dot_product_attention` | FP, INT | `FP32`, `INT8` | 8x8 |
 | `torch.nn.GELU` / `torch.nn.functional.gelu` | FP, INT | `FP32`, `BF16`, `INT8` | 8x8 |
 | `torch.nn.GroupNorm` / `torch.nn.functional.group_norm` | FP, INT | `FP32`, `INT8` | 8x8 |
 | `torch.nn.Hardsigmoid` / `torch.nn.functional.hardsigmoid` | FP, INT | `FP32`, `INT8` | 8x8 |