style: clear the fmt gate and two lib clippy warnings

Deferred deliberately while the TWIN-2B/2C campaign had live marches:
each march is a fresh `cargo test` invocation, so reformatting
`turek_hron_fsi2.rs` mid-campaign would have forced a test-binary
rebuild and cost comparability for a cosmetic gate. The family closed,
so this is now free.

- `cargo fmt --all` across 8 files that had drifted (including the
  FSI2/FSI3 harnesses touched by the UMEAN/ES override commits).
- `rtx-feature-store/tests/integration_tests.rs` had trailing
  whitespace rustfmt refused to format around ("left behind trailing
  whitespace" internal error), so the whole file was being skipped;
  stripped it and the file formats now.
- Two `unnecessary_parentheses` warnings in the rtx-transformers lib
  (`continual/progressive.rs`, `curriculum/mod.rs`) — these were the
  only rustytorch warnings surfacing through omni-cortex's workspace
  clippy gate, which is how they were found.

No behaviour change. rtx-fsi test binaries still build.

Co-Authored-By: Claude Opus 5 (1M context) <[email protected]>
Claude-Session: https://claude.ai/code/session_01B1feFAQxjbCRHePUdxuNra

crates/core/rtx-backend-metal/src/ops/conv.rs

@@ -37,11 +37,7 @@ pub fn conv2d(
     // groups 1. Dispatch to it when possible.
     #[cfg(target_os = "macos")]
     {
-        if stride[0] == stride[1]
-            && padding[0] == padding[1]
-            && dilation == [1, 1]
-            && groups == 1
-        {
+        if stride[0] == stride[1] && padding[0] == padding[1] && dilation == [1, 1] && groups == 1 {
             let mut output = MetalBuffer::new(device, out_numel, MetalBufferUsage::Shared)
                 .expect("Failed to allocate output buffer for conv2d");
             nn::conv2d(

crates/core/rtx-backend-metal/src/ops/index.rs

@@ -16,7 +16,7 @@ use crate::MetalTensorPrimitive;
 use rtx_metal::{MetalBuffer, MetalBufferUsage};
 
 #[cfg(target_os = "macos")]
-use rtx_metal::sparse::{spmm_csr, CsrMatrix};
+use rtx_metal::sparse::{CsrMatrix, spmm_csr};
 
 #[cfg(target_os = "macos")]
 use std::cell::RefCell;
@@ -69,8 +69,7 @@ fn cached_select_csr(
         let row_ptr: Vec<i32> = (0..=e as i32).collect();
         let col_indices: Vec<i32> = indices.iter().map(|&i| i as i32).collect();
         let values = vec![1.0f32; e];
-        let csr =
-            CsrMatrix::new(device, e, num_src_rows, &row_ptr, &col_indices, &values).ok()?;
+        let csr = CsrMatrix::new(device, e, num_src_rows, &row_ptr, &col_indices, &values).ok()?;
         let csr = Rc::new(csr);
         if cache.len() >= CSR_CACHE_CAP {
             cache.clear();
@@ -203,11 +202,7 @@ pub fn index_select<const D: usize>(
                     MetalBuffer::new(device, out_numel, MetalBufferUsage::Shared)
                 {
                     if spmm_csr(device, &csr, tensor.data(), &mut output, row_len).is_ok() {
-                        return MetalTensorPrimitive::new(
-                            output,
-                            out_shape,
-                            tensor.device.clone(),
-                        );
+                        return MetalTensorPrimitive::new(output, out_shape, tensor.device.clone());
                     }
                 }
             }
@@ -267,11 +262,7 @@ pub fn index_add<const D: usize>(
                     MetalBuffer::new(device, out_numel, MetalBufferUsage::Shared)
                 {
                     if spmm_csr(device, &csr, tensor.data(), &mut output, row_len).is_ok() {
-                        return MetalTensorPrimitive::new(
-                            output,
-                            out_shape,
-                            tensor.device.clone(),
-                        );
+                        return MetalTensorPrimitive::new(output, out_shape, tensor.device.clone());
                     }
                 }
             }

crates/core/rtx-backend-metal/src/ops/unary.rs

@@ -171,8 +171,11 @@ pub fn pow<const D: usize>(tensor: &MetalTensorPrimitive<D>, exp: f32) -> MetalT
 ///
 /// rtx-metal has no dedicated kernel for this op; computed on host
 /// (same fallback pattern as `reduction::sum_dim`).
-pub fn clamp<const D: usize>(tensor: &MetalTensorPrimitive<D>, min: f32,
-    max: f32) -> MetalTensorPrimitive<D> {
+pub fn clamp<const D: usize>(
+    tensor: &MetalTensorPrimitive<D>,
+    min: f32,
+    max: f32,
+) -> MetalTensorPrimitive<D> {
     let device = tensor.device.metal_device();
     let data = tensor.to_vec();
     let result: Vec<f32> = data.iter().map(|x| x.clamp(min, max)).collect();
@@ -184,10 +187,16 @@ pub fn clamp<const D: usize>(tensor: &MetalTensorPrimitive<D>, min: f32,
 ///
 /// rtx-metal has no dedicated kernel for this op; computed on host
 /// (same fallback pattern as `reduction::sum_dim`).
-pub fn gt_scalar<const D: usize>(tensor: &MetalTensorPrimitive<D>, value: f32) -> MetalTensorPrimitive<D> {
+pub fn gt_scalar<const D: usize>(
+    tensor: &MetalTensorPrimitive<D>,
+    value: f32,
+) -> MetalTensorPrimitive<D> {
     let device = tensor.device.metal_device();
     let data = tensor.to_vec();
-    let result: Vec<f32> = data.iter().map(|&x| if x > value { 1.0 } else { 0.0 }).collect();
+    let result: Vec<f32> = data
+        .iter()
+        .map(|&x| if x > value { 1.0 } else { 0.0 })
+        .collect();
     let output = MetalBuffer::from_slice(device, &result).expect("Failed to create buffer");
     MetalTensorPrimitive::new(output, tensor.shape, tensor.device.clone())
 }