diff --git a/.github/workflows/auto-tag.yml b/.github/workflows/auto-tag.yml
new file mode 100644
index 0000000..05b6962
--- /dev/null
+++ b/.github/workflows/auto-tag.yml
@@ -0,0 +1,22 @@
+name: Auto-tag
+
+on:
+  push:
+    branches: [main]
+
+jobs:
+  tag:
+    runs-on: ubuntu-latest
+    permissions:
+      contents: write
+    steps:
+      - uses: actions/checkout@v4
+      - name: Create tag if new version
+        run: |
+          VERSION=$(cargo metadata --no-deps --format-version 1 | jq -r '.packages[0].version')
+          if git ls-remote --tags origin "refs/tags/v$VERSION" | grep -q .; then
+            echo "Tag v$VERSION already exists, skipping"
+          else
+            git tag "v$VERSION"
+            git push origin "v$VERSION"
+          fi
diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.yml
index 994533a..184ecf1 100644
--- a/.github/workflows/ci.yml
+++ b/.github/workflows/ci.yml
@@ -32,8 +32,8 @@ jobs:
         with:
           toolchain: ${{ matrix.rust }}
       - uses: Swatinem/rust-cache@v2
-      - run: cargo test --all-features
-      - run: cargo test --doc --all-features
+      - run: cargo test --features std
+      - run: cargo test --doc --features std
 
   no-std:
     name: no_std
@@ -53,8 +53,8 @@ jobs:
           components: rustfmt, clippy
       - uses: Swatinem/rust-cache@v2
       - run: cargo fmt --all -- --check
-      - run: cargo clippy --all-targets --all-features -- -D warnings
-      - run: cargo doc --no-deps --all-features
+      - run: cargo clippy --all-targets --features std -- -D warnings
+      - run: cargo doc --no-deps --features std
         env:
           RUSTDOCFLAGS: -Dwarnings
 
@@ -65,7 +65,7 @@ jobs:
       - uses: actions/checkout@v4
       - uses: dtolnay/rust-toolchain@1.88
       - uses: Swatinem/rust-cache@v2
-      - run: cargo check --all-features
+      - run: cargo check --features std
 
   coverage:
     name: Coverage
@@ -75,7 +75,7 @@ jobs:
       - uses: dtolnay/rust-toolchain@stable
       - uses: Swatinem/rust-cache@v2
       - uses: taiki-e/install-action@cargo-tarpaulin
-      - run: cargo tarpaulin --out xml --all-features --exclude-files 'tools/*'
+      - run: cargo tarpaulin --out xml --features std --exclude-files 'tools/*'
       - uses: codecov/codecov-action@v4
         with:
           token: ${{ secrets.CODECOV_TOKEN }}
@@ -91,6 +91,29 @@ jobs:
         with:
           token: ${{ secrets.GITHUB_TOKEN }}
 
+  no-panic:
+    name: Verify No Panics
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      - uses: dtolnay/rust-toolchain@stable
+      - uses: Swatinem/rust-cache@v2
+
+      - name: Check all pub fns are annotated
+        run: |
+          pub_fns=$(grep -r '^pub fn ' src/ops/*.rs | wc -l)
+          annotations=$(grep -r 'no_panic::no_panic' src/ops/*.rs | wc -l)
+          echo "pub fn count: $pub_fns"
+          echo "no_panic annotations: $annotations"
+          if [ "$pub_fns" -ne "$annotations" ]; then
+            echo "::error::Not all public functions in src/ops/ have #[no_panic] annotation ($annotations/$pub_fns)"
+            grep -n '^pub fn ' src/ops/*.rs
+            exit 1
+          fi
+
+      - name: Build with no-panic verification
+        run: cargo build --profile no-panic-check --features verify-no-panic --bin verify_no_panic
+
   bench-check:
     name: Benchmarks Compile
     runs-on: ubuntu-latest
diff --git a/.github/workflows/publish.yml b/.github/workflows/publish.yml
index e7eae0b..2d936e0 100644
--- a/.github/workflows/publish.yml
+++ b/.github/workflows/publish.yml
@@ -17,9 +17,9 @@ jobs:
       - uses: actions/checkout@v4
       - uses: dtolnay/rust-toolchain@stable
       - uses: Swatinem/rust-cache@v2
-      - run: cargo test --all-features
+      - run: cargo test --features std
       - run: cargo test --no-default-features
-      - run: cargo doc --no-deps --all-features
+      - run: cargo doc --no-deps --features std
         env:
           RUSTDOCFLAGS: -Dwarnings
 
diff --git a/Cargo.toml b/Cargo.toml
index 48a95a7..48f38f1 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -1,6 +1,6 @@
 [package]
 name = "fixed_analytics"
-version = "0.5.1"
+version = "1.0.0"
 edition = "2024"
 rust-version = "1.88"
 authors = ["David Gathercole"]
@@ -13,28 +13,40 @@ keywords = ["cordic", "fixed-point", "trigonometry", "math", "no_std"]
 categories = ["mathematics", "no-std", "algorithms", "embedded"]
 
 [package.metadata.docs.rs]
-all-features = true
+features = ["std"]
 rustdoc-args = ["--cfg", "docsrs"]
 
 [features]
 default = ["std"]
 std = []
+verify-no-panic = ["dep:no-panic"]
 
 [dependencies]
-fixed = "1"
+fixed = "1.30"
+no-panic = { version = "0.1", optional = true }
 
 [dev-dependencies]
-criterion = { version = "0.8.1", features = ["html_reports"] }
+criterion = { version = "0.8", features = ["html_reports"] }
 
 [[bench]]
 name = "benchmarks"
 harness = false
 
+[[bin]]
+name = "verify_no_panic"
+required-features = ["verify-no-panic"]
+
 [profile.release]
 lto = true
 codegen-units = 1
 panic = "abort"
 
+[profile.no-panic-check]
+inherits = "release"
+lto = "fat"
+codegen-units = 1
+panic = "unwind"
+
 [profile.bench]
 lto = true
 codegen-units = 1
diff --git a/README.md b/README.md
index 0ff024a..2586f9f 100644
--- a/README.md
+++ b/README.md
@@ -1,6 +1,6 @@
 # fixed_analytics
 
-Fixed-point mathematical functions which are accurate, fast, safe, and machine independent.
+Fixed-point mathematical functions which are accurate, deterministic, and guaranteed not to panic.
 
 [![Crates.io](https://img.shields.io/crates/v/fixed_analytics.svg)](https://crates.io/crates/fixed_analytics)
 [![CI](https://github.com/GeEom/fixed_analytics/actions/workflows/ci.yml/badge.svg)](https://github.com/GeEom/fixed_analytics/actions/workflows/ci.yml)
@@ -30,14 +30,14 @@ Requires Rust 1.88 or later.
 
 ```toml
 [dependencies]
-fixed_analytics = "0.5.1"
+fixed_analytics = "1.0.0"
 ```
 
 For `no_std` environments:
 
 ```toml
 [dependencies]
-fixed_analytics = { version = "0.5.1", default-features = false }
+fixed_analytics = { version = "1.0.0", default-features = false }
 ```
 
 ## Available Functions
@@ -54,7 +54,7 @@ fixed_analytics = { version = "0.5.1", default-features = false }
 | Exponential | `exp`, `pow2` | `ln`, `log2`, `log10` |
 | Algebraic | — | `sqrt` |
 
-Functions are calculated via CORDIC, Newton-Raphson, and Taylor series techniques.
+Functions are calculated via CORDIC, Newton-Raphson, and Taylor series techniques. Complete absence of panic is verified at the linker level via the [`no-panic`](https://github.com/dtolnay/no-panic) crate.
 
 ### Saturation Behavior
 
@@ -75,7 +75,7 @@ Where for `tan`, "pole" refers to ±π/2, ±3π/2, ±5π/2, ...
 <!-- ACCURACY_START -->
 ### Accuracy
 
-Relative error statistics measured against MPFR reference implementations. The file tools/accuracy-bench/baseline.json contains further measurements.
+Relative error statistics measured against MPFR reference implementations. Accuracy regressions are not permitted; every change is benchmarked against the baseline before merging. The file tools/accuracy-bench/baseline.json contains further measurements.
 
 | Function | I16F16 Mean | I16F16 Median | I16F16 P95 | I32F32 Mean | I32F32 Median | I32F32 P95 |
 |----------|-------------|---------------|------------|-------------|---------------|------------|
diff --git a/src/bin/verify_no_panic.rs b/src/bin/verify_no_panic.rs
new file mode 100644
index 0000000..6d1c735
--- /dev/null
+++ b/src/bin/verify_no_panic.rs
@@ -0,0 +1,56 @@
+//! Binary that instantiates every public function with a concrete type.
+//!
+//! This exists solely to trigger monomorphization so that `no_panic`'s
+//! linker-level check can verify that no panic paths survive optimization.
+//! It is only compiled under the `verify-no-panic` feature.
+
+#[cfg(not(feature = "verify-no-panic"))]
+compile_error!("this binary should only be built with --features verify-no-panic");
+
+use fixed::types::I16F16;
+use fixed_analytics::bounded::{NonNegative, OpenUnitInterval};
+use fixed_analytics::ops::algebraic::sqrt_nonneg;
+use fixed_analytics::ops::hyperbolic::atanh_open;
+use fixed_analytics::{
+    acos, acosh, acoth, asin, asinh, atan, atan2, atanh, cos, cosh, coth, exp, ln, log2, log10,
+    pow2, sin, sin_cos, sinh, sinh_cosh, sqrt, tan, tanh,
+};
+
+fn main() {
+    // Use black_box to prevent the optimizer from eliminating calls entirely.
+    let x = std::hint::black_box(I16F16::from_num(0.5));
+    let y = std::hint::black_box(I16F16::from_num(0.25));
+
+    // Total functions (return T)
+    let _ = std::hint::black_box(sin(x));
+    let _ = std::hint::black_box(cos(x));
+    let _ = std::hint::black_box(tan(x));
+    let _ = std::hint::black_box(sin_cos(x));
+    let _ = std::hint::black_box(atan(x));
+    let _ = std::hint::black_box(atan2(y, x));
+    let _ = std::hint::black_box(exp(x));
+    let _ = std::hint::black_box(pow2(x));
+    let _ = std::hint::black_box(sinh(x));
+    let _ = std::hint::black_box(cosh(x));
+    let _ = std::hint::black_box(tanh(x));
+    let _ = std::hint::black_box(sinh_cosh(x));
+    let _ = std::hint::black_box(asinh(x));
+
+    // Fallible functions (return Result<T>)
+    let _ = std::hint::black_box(asin(x));
+    let _ = std::hint::black_box(acos(x));
+    let _ = std::hint::black_box(sqrt(x));
+    let _ = std::hint::black_box(ln(x));
+    let _ = std::hint::black_box(log2(x));
+    let _ = std::hint::black_box(log10(x));
+    let _ = std::hint::black_box(acosh(I16F16::from_num(2)));
+    let _ = std::hint::black_box(atanh(x));
+    let _ = std::hint::black_box(coth(x));
+    let _ = std::hint::black_box(acoth(I16F16::from_num(2)));
+
+    // Type-safe wrapper functions
+    let nn = NonNegative::new(x).unwrap();
+    let _ = std::hint::black_box(sqrt_nonneg(nn));
+    let ou = OpenUnitInterval::new(x).unwrap();
+    let _ = std::hint::black_box(atanh_open(ou));
+}
diff --git a/src/ops/algebraic.rs b/src/ops/algebraic.rs
index 51c5534..2647aeb 100644
--- a/src/ops/algebraic.rs
+++ b/src/ops/algebraic.rs
@@ -9,6 +9,7 @@ use crate::traits::CordicNumber;
 /// # Errors
 /// Returns `DomainError` if `x < 0`.
 #[must_use = "returns the square root result which should be handled"]
+#[cfg_attr(feature = "verify-no-panic", no_panic::no_panic)]
 pub fn sqrt<T: CordicNumber>(x: T) -> Result<T> {
     NonNegative::new(x)
         .map(sqrt_nonneg)
@@ -23,6 +24,7 @@ pub fn sqrt<T: CordicNumber>(x: T) -> Result<T> {
 /// Use this when the non-negativity of the input is already established
 /// through mathematical invariants (e.g., `1 + x²`, `1 - x²` for `|x| ≤ 1`).
 #[must_use]
+#[cfg_attr(feature = "verify-no-panic", no_panic::no_panic)]
 pub fn sqrt_nonneg<T: CordicNumber>(x: NonNegative<T>) -> T {
     let x = x.get();
     let zero = T::zero();
@@ -85,9 +87,9 @@ pub fn sqrt_nonneg<T: CordicNumber>(x: NonNegative<T>) -> T {
         let new_guess = sum.saturating_mul(half);
 
         let diff = if new_guess > guess {
-            new_guess - guess
+            new_guess.saturating_sub(guess)
         } else {
-            guess - new_guess
+            guess.saturating_sub(new_guess)
         };
 
         if diff <= epsilon {
diff --git a/src/ops/circular.rs b/src/ops/circular.rs
index 87d5e46..a8c5b76 100644
--- a/src/ops/circular.rs
+++ b/src/ops/circular.rs
@@ -8,6 +8,7 @@ use crate::traits::CordicNumber;
 
 /// Sine and cosine. More efficient than separate calls. Accepts any angle.
 #[must_use]
+#[cfg_attr(feature = "verify-no-panic", no_panic::no_panic)]
 pub fn sin_cos<T: CordicNumber>(angle: T) -> (T, T) {
     let pi = T::pi();
     let frac_pi_2 = T::frac_pi_2();
@@ -58,6 +59,7 @@ pub fn sin_cos<T: CordicNumber>(angle: T) -> (T, T) {
 /// Sine. Accepts any angle (reduced internally).
 #[inline]
 #[must_use]
+#[cfg_attr(feature = "verify-no-panic", no_panic::no_panic)]
 pub fn sin<T: CordicNumber>(angle: T) -> T {
     sin_cos(angle).0
 }
@@ -65,6 +67,7 @@ pub fn sin<T: CordicNumber>(angle: T) -> T {
 /// Cosine. Accepts any angle (reduced internally).
 #[inline]
 #[must_use]
+#[cfg_attr(feature = "verify-no-panic", no_panic::no_panic)]
 pub fn cos<T: CordicNumber>(angle: T) -> T {
     sin_cos(angle).1
 }
@@ -108,6 +111,7 @@ pub fn cos<T: CordicNumber>(angle: T) -> T {
 /// }
 /// ```
 #[must_use]
+#[cfg_attr(feature = "verify-no-panic", no_panic::no_panic)]
 pub fn tan<T: CordicNumber>(angle: T) -> T {
     let (s, c) = sin_cos(angle);
     s.div(c)
@@ -118,6 +122,7 @@ pub fn tan<T: CordicNumber>(angle: T) -> T {
 /// # Errors
 /// Returns `DomainError` if `|x| > 1`.
 #[must_use = "returns the arcsine result which should be handled"]
+#[cfg_attr(feature = "verify-no-panic", no_panic::no_panic)]
 pub fn asin<T: CordicNumber>(x: T) -> Result<T> {
     let Some(unit_x) = UnitInterval::new(x) else {
         return Err(Error::domain("asin", "value in range [-1, 1]"));
@@ -156,13 +161,15 @@ pub fn asin<T: CordicNumber>(x: T) -> Result<T> {
 /// # Errors
 /// Returns `DomainError` if `|x| > 1`.
 #[must_use = "returns the arccosine result which should be handled"]
+#[cfg_attr(feature = "verify-no-panic", no_panic::no_panic)]
 pub fn acos<T: CordicNumber>(x: T) -> Result<T> {
     // acos(x) = π/2 - asin(x)
-    asin(x).map(|a| T::frac_pi_2() - a)
+    asin(x).map(|a| T::frac_pi_2().saturating_sub(a))
 }
 
 /// Arctangent. Accepts any value. Returns angle in `(-π/2, π/2)`.
 #[must_use]
+#[cfg_attr(feature = "verify-no-panic", no_panic::no_panic)]
 pub fn atan<T: CordicNumber>(x: T) -> T {
     let zero = T::zero();
     let one = T::one();
@@ -192,6 +199,7 @@ pub fn atan<T: CordicNumber>(x: T) -> T {
 
 /// Four-quadrant arctangent. Returns angle in `[-π, π]`. Returns 0 for (0, 0).
 #[must_use]
+#[cfg_attr(feature = "verify-no-panic", no_panic::no_panic)]
 pub fn atan2<T: CordicNumber>(y: T, x: T) -> T {
     let zero = T::zero();
     let pi = T::pi();
diff --git a/src/ops/exponential.rs b/src/ops/exponential.rs
index e2330c7..47655fe 100644
--- a/src/ops/exponential.rs
+++ b/src/ops/exponential.rs
@@ -37,6 +37,7 @@ use crate::traits::CordicNumber;
 /// }
 /// ```
 #[must_use]
+#[cfg_attr(feature = "verify-no-panic", no_panic::no_panic)]
 pub fn exp<T: CordicNumber>(x: T) -> T {
     let zero = T::zero();
     let one = T::one();
@@ -55,7 +56,7 @@ pub fn exp<T: CordicNumber>(x: T) -> T {
     // Reduce positive values
     let mut i = 0;
     while reduced > ln2 && i < 64 {
-        reduced -= ln2;
+        reduced = reduced.saturating_sub(ln2);
         scale += 1;
         i += 1;
     }
@@ -63,7 +64,7 @@ pub fn exp<T: CordicNumber>(x: T) -> T {
     // Reduce negative values
     i = 0;
     while reduced < -ln2 && i < 64 {
-        reduced += ln2;
+        reduced = reduced.saturating_add(ln2);
         scale -= 1;
         i += 1;
     }
@@ -100,6 +101,7 @@ pub fn exp<T: CordicNumber>(x: T) -> T {
 /// # Errors
 /// Returns `DomainError` if `x ≤ 0`.
 #[must_use = "returns the natural logarithm result which should be handled"]
+#[cfg_attr(feature = "verify-no-panic", no_panic::no_panic)]
 pub fn ln<T: CordicNumber>(x: T) -> Result<T> {
     let zero = T::zero();
     let one = T::one();
@@ -128,15 +130,15 @@ pub fn ln<T: CordicNumber>(x: T) -> Result<T> {
     let mut i = 0;
     while normalized > two && i < 128 {
         normalized = normalized >> 1;
-        k_ln2 += ln2;
+        k_ln2 = k_ln2.saturating_add(ln2);
         i += 1;
     }
 
     // For small x (< 0.5), multiply by 2 repeatedly
     i = 0;
     while normalized < half && i < 128 {
-        normalized = normalized + normalized;
-        k_ln2 -= ln2;
+        normalized = normalized.saturating_add(normalized);
+        k_ln2 = k_ln2.saturating_sub(ln2);
         i += 1;
     }
 
@@ -150,9 +152,9 @@ pub fn ln<T: CordicNumber>(x: T) -> Result<T> {
     let arg = OpenUnitInterval::from_normalized_ln_arg(norm);
 
     let atanh_val = atanh_open(arg);
-    let ln_normalized = atanh_val + atanh_val; // 2 * atanh
+    let ln_normalized = atanh_val.saturating_add(atanh_val); // 2 * atanh
 
-    Ok(ln_normalized + k_ln2)
+    Ok(ln_normalized.saturating_add(k_ln2))
 }
 
 /// Base-2 logarithm. Domain: `x > 0`.
@@ -160,6 +162,7 @@ pub fn ln<T: CordicNumber>(x: T) -> Result<T> {
 /// # Errors
 /// Returns `DomainError` if `x ≤ 0`.
 #[must_use = "returns the base-2 logarithm result which should be handled"]
+#[cfg_attr(feature = "verify-no-panic", no_panic::no_panic)]
 pub fn log2<T: CordicNumber>(x: T) -> Result<T> {
     let ln_x = ln(x)?;
     let ln_2 = T::ln_2();
@@ -171,6 +174,7 @@ pub fn log2<T: CordicNumber>(x: T) -> Result<T> {
 /// # Errors
 /// Returns `DomainError` if `x ≤ 0`.
 #[must_use = "returns the base-10 logarithm result which should be handled"]
+#[cfg_attr(feature = "verify-no-panic", no_panic::no_panic)]
 pub fn log10<T: CordicNumber>(x: T) -> Result<T> {
     let ln_x = ln(x)?;
     let ln_10 = T::ln_10();
@@ -192,6 +196,7 @@ pub fn log10<T: CordicNumber>(x: T) -> Result<T> {
 /// - **I16F16:** Saturates for x > ~15 or x < ~-16
 /// - **I32F32:** Saturates for x > ~31 or x < ~-32
 #[must_use]
+#[cfg_attr(feature = "verify-no-panic", no_panic::no_panic)]
 pub fn pow2<T: CordicNumber>(x: T) -> T {
     let ln_2 = T::ln_2();
     exp(x.saturating_mul(ln_2))
diff --git a/src/ops/hyperbolic.rs b/src/ops/hyperbolic.rs
index 9155e51..788e47e 100644
--- a/src/ops/hyperbolic.rs
+++ b/src/ops/hyperbolic.rs
@@ -58,27 +58,25 @@ const ATANH_REDUCTION_THRESHOLD_I1F63: i64 = 0x6000_0000_0000_0000;
 /// at the same rate), so the relationship cosh²(x) - sinh²(x) = 1
 /// will not hold for saturated outputs.
 #[must_use]
+#[cfg_attr(feature = "verify-no-panic", no_panic::no_panic)]
 pub fn sinh_cosh<T: CordicNumber>(x: T) -> (T, T) {
     let zero = T::zero();
     let one = T::one();
     // Compute limit as 1 + fractional_part (~1.1182)
     let limit = one.saturating_add(T::from_i1f63(HYPERBOLIC_CONVERGENCE_LIMIT_FRAC_I1F63));
 
-    // Handle argument reduction for large values
-    if x.abs() > limit {
-        // Use the identities:
-        // sinh(2x) = 2 * sinh(x) * cosh(x)
-        // cosh(2x) = cosh²(x) + sinh²(x)
-        let half_x = x >> 1;
-        let (sh, ch) = sinh_cosh(half_x);
-
-        let sinh_result = sh.saturating_mul(ch).saturating_mul(T::two());
-        let cosh_result = ch.saturating_mul(ch).saturating_add(sh.saturating_mul(sh));
-
-        return (sinh_result, cosh_result);
+    // Iterative argument reduction for large values.
+    // Count how many halvings are needed, then rebuild on the way back.
+    let mut reduced = x;
+    let mut depth: u32 = 0;
+    // Max depth bounded by bit width: each halving shifts right by 1,
+    // so after total_bits iterations the value is zero and below the limit.
+    while reduced.abs() > limit && depth < T::total_bits() {
+        reduced = reduced >> 1;
+        depth += 1;
     }
 
-    // For very small x, use Taylor series approximation to avoid CORDIC
+    // For very small reduced, use Taylor series approximation to avoid CORDIC
     // overshoot on the first iteration (where atanh(0.5) ≈ 0.549 is larger than x).
     // Use precision-dependent threshold and order.
     let threshold_bits = if T::frac_bits() >= 24 {
@@ -87,41 +85,47 @@ pub fn sinh_cosh<T: CordicNumber>(x: T) -> (T, T) {
         TAYLOR_THRESHOLD_LOW_PREC_I1F63
     };
     let small_threshold = T::from_i1f63(threshold_bits);
-    if x.abs() < small_threshold {
-        let x_sq = x.saturating_mul(x);
-        let x_cu = x_sq.saturating_mul(x);
+    let (mut sh, mut ch) = if reduced.abs() < small_threshold {
+        let x_sq = reduced.saturating_mul(reduced);
+        let x_cu = x_sq.saturating_mul(reduced);
         let x_qu = x_sq.saturating_mul(x_sq);
 
         // Higher precision benefits from higher-order Taylor terms
         if T::frac_bits() >= 24 {
             // sinh(x) ≈ x + x³/6 + x⁵/120, cosh(x) ≈ 1 + x²/2 + x⁴/24 + x⁶/720
-            let x_5 = x_qu.saturating_mul(x);
+            let x_5 = x_qu.saturating_mul(reduced);
             let x_6 = x_qu.saturating_mul(x_sq);
-            // cosh base: 1 + x²/2 + x⁴/24
             let cosh_base = one
                 .saturating_add(x_sq >> 1)
                 .saturating_add(x_qu.div(T::from_num(24)));
             let cosh_approx = cosh_base.saturating_add(x_6.div(T::from_num(720)));
-            // sinh base: x + x³/6
-            let sinh_base = x.saturating_add(x_cu.div(T::from_num(6)));
+            let sinh_base = reduced.saturating_add(x_cu.div(T::from_num(6)));
             let sinh_approx = sinh_base.saturating_add(x_5.div(T::from_num(120)));
-            return (sinh_approx, cosh_approx);
+            (sinh_approx, cosh_approx)
+        } else {
+            // sinh(x) ≈ x + x³/6, cosh(x) ≈ 1 + x²/2 + x⁴/24
+            let c = one.saturating_add(x_sq >> 1);
+            let cosh_approx = c.saturating_add(x_qu.div(T::from_num(24)));
+            let sinh_approx = reduced.saturating_add(x_cu.div(T::from_num(6)));
+            (sinh_approx, cosh_approx)
         }
-        // sinh(x) ≈ x + x³/6, cosh(x) ≈ 1 + x²/2 + x⁴/24
-        let c = one.saturating_add(x_sq >> 1);
-        let cosh_approx = c.saturating_add(x_qu.div(T::from_num(24)));
-        let sinh_approx = x.saturating_add(x_cu.div(T::from_num(6)));
-        return (sinh_approx, cosh_approx);
-    }
-
-    // For moderate x, use CORDIC directly.
-    // Hyperbolic CORDIC scales results by 1/K_h ≈ 1.2075.
-    // To compensate, we pre-multiply by 1/K_h (using precomputed constant).
-    let inv_gain = hyperbolic_gain_inv(); // 1/K_h ≈ 1.2075
+    } else {
+        // For moderate x, use CORDIC directly.
+        let inv_gain = hyperbolic_gain_inv();
+        let (cosh_val, sinh_val, _) = hyperbolic_rotation(inv_gain, zero, reduced);
+        (sinh_val, cosh_val)
+    };
 
-    let (cosh_val, sinh_val, _) = hyperbolic_rotation(inv_gain, zero, x);
+    // Reconstruct via doubling: sinh(2x) = 2·sinh(x)·cosh(x),
+    //                           cosh(2x) = cosh²(x) + sinh²(x)
+    for _ in 0..depth {
+        let new_sh = sh.saturating_mul(ch).saturating_mul(T::two());
+        let new_ch = ch.saturating_mul(ch).saturating_add(sh.saturating_mul(sh));
+        sh = new_sh;
+        ch = new_ch;
+    }
 
-    (sinh_val, cosh_val)
+    (sh, ch)
 }
 
 /// Hyperbolic sine.
@@ -144,6 +148,7 @@ pub fn sinh_cosh<T: CordicNumber>(x: T) -> (T, T) {
 /// with argument reduction for |x| > 1.118.
 #[inline]
 #[must_use]
+#[cfg_attr(feature = "verify-no-panic", no_panic::no_panic)]
 pub fn sinh<T: CordicNumber>(x: T) -> T {
     sinh_cosh(x).0
 }
@@ -166,12 +171,14 @@ pub fn sinh<T: CordicNumber>(x: T) -> T {
 /// - **I32F32:** Saturates for |x| > ~21.5
 #[inline]
 #[must_use]
+#[cfg_attr(feature = "verify-no-panic", no_panic::no_panic)]
 pub fn cosh<T: CordicNumber>(x: T) -> T {
     sinh_cosh(x).1
 }
 
 /// Hyperbolic tangent. Result in `(-1, 1)`.
 #[must_use]
+#[cfg_attr(feature = "verify-no-panic", no_panic::no_panic)]
 pub fn tanh<T: CordicNumber>(x: T) -> T {
     let (s, c) = sinh_cosh(x);
     s.div(c)
@@ -182,6 +189,7 @@ pub fn tanh<T: CordicNumber>(x: T) -> T {
 /// # Errors
 /// Returns `DomainError` if `x = 0`.
 #[must_use = "returns the hyperbolic cotangent result which should be handled"]
+#[cfg_attr(feature = "verify-no-panic", no_panic::no_panic)]
 pub fn coth<T: CordicNumber>(x: T) -> Result<T> {
     if x == T::zero() {
         return Err(Error::domain("coth", "non-zero value"));
@@ -192,6 +200,7 @@ pub fn coth<T: CordicNumber>(x: T) -> Result<T> {
 
 /// Inverse hyperbolic sine. Accepts any value.
 #[must_use]
+#[cfg_attr(feature = "verify-no-panic", no_panic::no_panic)]
 pub fn asinh<T: CordicNumber>(x: T) -> T {
     if x == T::zero() {
         return T::zero();
@@ -212,6 +221,7 @@ pub fn asinh<T: CordicNumber>(x: T) -> T {
 /// # Errors
 /// Returns `DomainError` if `x < 1`.
 #[must_use = "returns the inverse hyperbolic cosine result which should be handled"]
+#[cfg_attr(feature = "verify-no-panic", no_panic::no_panic)]
 pub fn acosh<T: CordicNumber>(x: T) -> Result<T> {
     let at_least_one = AtLeastOne::new(x).ok_or_else(|| Error::domain("acosh", "value >= 1"))?;
 
@@ -234,6 +244,7 @@ pub fn acosh<T: CordicNumber>(x: T) -> Result<T> {
 /// # Errors
 /// Returns `DomainError` if `|x| ≥ 1`.
 #[must_use = "returns the inverse hyperbolic tangent result which should be handled"]
+#[cfg_attr(feature = "verify-no-panic", no_panic::no_panic)]
 pub fn atanh<T: CordicNumber>(x: T) -> Result<T> {
     OpenUnitInterval::new(x)
         .map(atanh_open)
@@ -248,6 +259,7 @@ pub fn atanh<T: CordicNumber>(x: T) -> Result<T> {
 /// Use this when the input is known to be in (-1, 1) through mathematical
 /// invariants (e.g., `x / sqrt(1 + x²)`).
 #[must_use]
+#[cfg_attr(feature = "verify-no-panic", no_panic::no_panic)]
 pub fn atanh_open<T: CordicNumber>(x: OpenUnitInterval<T>) -> T {
     atanh_core(x.get())
 }
@@ -263,31 +275,36 @@ fn atanh_core<T: CordicNumber>(x: T) -> T {
 
     let threshold = T::from_i1f63(ATANH_REDUCTION_THRESHOLD_I1F63);
 
+    // Fast path: no argument reduction needed, use CORDIC directly.
     if x.abs() <= threshold {
-        // Direct CORDIC computation
         let (_, _, z) = hyperbolic_vectoring(one, x, zero);
         return z;
     }
 
-    // Argument reduction using the identity:
-    // atanh(x) = atanh(a) + atanh((x - a) / (1 - a*x))
-    // We use a = 0.5, for which atanh(0.5) is a precomputed constant.
+    // Argument reduction needed. Work with |x| and track sign.
     let half = T::half();
     let atanh_half = T::from_i1f63(crate::tables::hyperbolic::ATANH_HALF);
-
     let sign = if x.is_negative() { -one } else { one };
-    let abs_x = x.abs();
-
-    // Compute reduced argument: (|x| - 0.5) / (1 - 0.5*|x|)
-    let numerator = abs_x - half;
-    let denominator = one - half.saturating_mul(abs_x);
-    let reduced = numerator.div(denominator);
+    let mut abs_x = x.abs();
+
+    // Iterative argument reduction: each step reduces |x| and accumulates
+    // atanh(0.5) into the result. Max iterations bounded by frac_bits since
+    // each reduction roughly halves the distance from the threshold.
+    let mut accumulated = zero;
+    let mut i: u32 = 0;
+    while abs_x > threshold && i < T::frac_bits() {
+        // atanh(x) = atanh(0.5) + atanh((x - 0.5) / (1 - 0.5*x))
+        let numerator = abs_x.saturating_sub(half);
+        let denominator = one.saturating_sub(half.saturating_mul(abs_x));
+        abs_x = numerator.div(denominator);
+        accumulated = accumulated.saturating_add(atanh_half);
+        i += 1;
+    }
 
-    // Recursively compute atanh of reduced argument
-    let atanh_reduced = atanh_core(reduced);
+    // Direct CORDIC computation on the reduced argument
+    let (_, _, z) = hyperbolic_vectoring(one, abs_x, zero);
 
-    // atanh(x) = sign * (atanh(0.5) + atanh(reduced))
-    sign.saturating_mul(atanh_half.saturating_add(atanh_reduced))
+    sign.saturating_mul(accumulated.saturating_add(z))
 }
 
 /// Inverse hyperbolic cotangent. Domain: `|x| > 1`.
@@ -295,6 +312,7 @@ fn atanh_core<T: CordicNumber>(x: T) -> T {
 /// # Errors
 /// Returns `DomainError` if `|x| ≤ 1`.
 #[must_use = "returns the inverse hyperbolic cotangent result which should be handled"]
+#[cfg_attr(feature = "verify-no-panic", no_panic::no_panic)]
 pub fn acoth<T: CordicNumber>(x: T) -> Result<T> {
     let one = T::one();
 
diff --git a/src/traits.rs b/src/traits.rs
index 48dd169..0bbe64f 100644
--- a/src/traits.rs
+++ b/src/traits.rs
@@ -1,7 +1,7 @@
 //! Trait definitions for types compatible with CORDIC algorithms.
 
 use core::ops::{Add, AddAssign, Mul, Neg, Shl, Shr, Sub, SubAssign};
-use fixed::traits::Fixed;
+use fixed::traits::{Fixed, FixedSigned};
 use fixed::types::extra::{IsLessOrEqual, LeEqU128, True, Unsigned};
 use fixed::{FixedI8, FixedI16, FixedI32, FixedI64, FixedI128};
 
@@ -177,7 +177,7 @@ macro_rules! impl_cordic_generic {
 
             #[inline]
             fn abs(self) -> Self {
-                if self.is_negative() { -self } else { self }
+                FixedSigned::saturating_abs(self)
             }
 
             #[inline]
@@ -276,7 +276,17 @@ macro_rules! impl_cordic_generic {
 
             #[inline]
             fn div(self, rhs: Self) -> Self {
-                self / rhs
+                match Fixed::checked_div(self, rhs) {
+                    Some(v) => v,
+                    // Division by zero or overflow: saturate based on sign agreement.
+                    None => {
+                        if self.is_negative() != rhs.is_negative() {
+                            Self::MIN
+                        } else {
+                            Self::MAX
+                        }
+                    }
+                }
             }
 
             #[inline]
diff --git a/tools/accuracy-bench/Cargo.toml b/tools/accuracy-bench/Cargo.toml
index d31456a..48f753e 100644
--- a/tools/accuracy-bench/Cargo.toml
+++ b/tools/accuracy-bench/Cargo.toml
@@ -6,9 +6,9 @@ publish = false
 
 [dependencies]
 fixed_analytics = { path = "../.." }
-fixed = "1"
-rug = "1.27"
-serde = { version = "1", features = ["derive"] }
-serde_json = "1"
-comfy-table = "7"
+fixed = "1.30"
+rug = "1.28"
+serde = { version = "1.0", features = ["derive"] }
+serde_json = "1.0"
+comfy-table = "7.2"
 rayon = "1.10"
diff --git a/tools/accuracy-bench/baseline.json b/tools/accuracy-bench/baseline.json
index a6ee989..62e77a7 100644
--- a/tools/accuracy-bench/baseline.json
+++ b/tools/accuracy-bench/baseline.json
@@ -1,5 +1,5 @@
 {
-  "timestamp": 1767828080,
+  "timestamp": 1773659600,
   "results": [
     {
       "name": "sin",
@@ -186,28 +186,28 @@
       "i16f16": {
         "count": 59007,
         "abs_max": 1.0772810843461684,
-        "abs_mean": 0.06303729385570156,
+        "abs_mean": 0.06303677330481015,
         "abs_p50": 0.004098231957545551,
         "abs_p95": 0.34691838864932834,
         "abs_p99": 0.6188876148160034,
         "rel_max": 0.0049975358676744045,
-        "rel_mean": 0.00018543208668244187,
-        "rel_p50": 0.00013881257641081612,
-        "rel_p95": 0.0005054464161457844,
-        "rel_p99": 0.000659939413180731
+        "rel_mean": 0.00018236789394986054,
+        "rel_p50": 0.0001344182189823812,
+        "rel_p95": 0.0005027598796161228,
+        "rel_p99": 0.0006566622127170074
       },
       "i32f32": {
         "count": 59007,
         "abs_max": 0.008872128233271326,
-        "abs_mean": 1.396564603408181e-6,
+        "abs_mean": 1.396558004830551e-6,
         "abs_p50": 7.002165069991406e-8,
         "abs_p95": 5.8817661283683265e-6,
         "abs_p99": 0.000011134297892567702,
         "rel_max": 0.00020256339056303632,
-        "rel_mean": 1.058698242414027e-8,
-        "rel_p50": 2.3731117361650507e-9,
-        "rel_p95": 9.576798998259808e-9,
-        "rel_p99": 1.3814111163622916e-8
+        "rel_mean": 1.0480732995514983e-8,
+        "rel_p50": 2.3544956150429304e-9,
+        "rel_p95": 9.303074988477954e-9,
+        "rel_p99": 1.285273873766771e-8
       },
       "samples_tested": 59007
     },
@@ -216,12 +216,12 @@
       "i16f16": {
         "count": 59007,
         "abs_max": 1.0772732080154128,
-        "abs_mean": 0.0630310547325392,
+        "abs_mean": 0.06303092881225811,
         "abs_p50": 0.004094172796769158,
         "abs_p95": 0.3469277889677187,
         "abs_p99": 0.6188732979264842,
         "rel_max": 0.0010691901922094571,
-        "rel_mean": 0.00017297585484030712,
+        "rel_mean": 0.00017285180819279924,
         "rel_p50": 0.00012268943900841178,
         "rel_p95": 0.0005004633648159678,
         "rel_p99": 0.0006510488630542412
@@ -229,12 +229,12 @@
       "i32f32": {
         "count": 59007,
         "abs_max": 0.008869816353453075,
-        "abs_mean": 1.39642538721307e-6,
+        "abs_mean": 1.3964231030204069e-6,
         "abs_p50": 7.00661288988158e-8,
         "abs_p95": 5.882438131266099e-6,
         "abs_p99": 0.000011134392480016686,
         "rel_max": 0.00020245784587786796,
-        "rel_mean": 1.0250328326619625e-8,
+        "rel_mean": 1.0248053709258805e-8,
         "rel_p50": 2.073955819960864e-9,
         "rel_p95": 9.159278823861197e-9,
         "rel_p99": 1.2647926134213153e-8
@@ -246,28 +246,28 @@
       "i16f16": {
         "count": 59007,
         "abs_max": 0.0001416839219182675,
-        "abs_mean": 0.000013029573082792987,
-        "abs_p50": 0.000013432866548424016,
-        "abs_p95": 0.00003051174541490731,
-        "abs_p99": 0.00006235363517403947,
+        "abs_mean": 0.000012729098112476967,
+        "abs_p50": 0.000013373680597172921,
+        "abs_p95": 0.000027615223236021613,
+        "abs_p99": 0.000057425093706209296,
         "rel_max": 0.004137583155192043,
-        "rel_mean": 0.000022637207713695504,
-        "rel_p50": 0.00001383955161030536,
-        "rel_p95": 0.00006125969525471107,
-        "rel_p99": 0.0002692388398758907
+        "rel_mean": 0.000020784111846600086,
+        "rel_p50": 0.0000138370779552808,
+        "rel_p95": 0.000058866070650500954,
+        "rel_p99": 0.00018729767889252025
       },
       "i32f32": {
         "count": 59007,
         "abs_max": 0.000020695083288946314,
-        "abs_mean": 1.2095156976429015e-9,
-        "abs_p50": 1.2857270803579013e-10,
-        "abs_p95": 6.234735505650235e-10,
-        "abs_p99": 1.298984919628765e-9,
+        "abs_mean": 1.2050832301560682e-9,
+        "abs_p50": 1.2859258102793092e-10,
+        "abs_p95": 5.812544889849391e-10,
+        "abs_p99": 1.2296446916248982e-9,
         "rel_max": 0.00002564595268436019,
-        "rel_mean": 1.7074502719755845e-9,
-        "rel_p50": 1.312886102619071e-10,
-        "rel_p95": 1.220758508492699e-9,
-        "rel_p99": 6.378228306145412e-9
+        "rel_mean": 1.6354159982679383e-9,
+        "rel_p50": 1.3132154604890873e-10,
+        "rel_p95": 1.2321919801854915e-9,
+        "rel_p99": 4.89240139683993e-9
       },
       "samples_tested": 59007
     },
@@ -275,29 +275,29 @@
       "name": "coth",
       "i16f16": {
         "count": 59003,
-        "abs_max": 0.009257363297408006,
-        "abs_mean": 0.00006385935542136119,
-        "abs_p50": 4.967882137663082e-6,
-        "abs_p95": 0.0001682281371442329,
-        "abs_p99": 0.0017492381017776282,
-        "rel_max": 0.0009808148379961716,
-        "rel_mean": 0.000017359339848271247,
-        "rel_p50": 4.861209237654236e-6,
-        "rel_p95": 0.00007231630780889138,
-        "rel_p99": 0.00027457495901270164
+        "abs_max": 0.00186560782194789,
+        "abs_mean": 0.000027539381408770688,
+        "abs_p50": 4.961126731650722e-6,
+        "abs_p95": 0.00013365012785548913,
+        "abs_p99": 0.00048479706134862965,
+        "rel_max": 0.0003172539281558055,
+        "rel_mean": 0.000011996517339386098,
+        "rel_p50": 4.82788872128744e-6,
+        "rel_p95": 0.00005573355105518167,
+        "rel_p99": 0.00011790663206009863
       },
       "i32f32": {
         "count": 59003,
         "abs_max": 5.297342871246613e-6,
-        "abs_mean": 1.320947570471908e-9,
-        "abs_p50": 1.4142131909977707e-10,
-        "abs_p95": 3.025110828502875e-9,
-        "abs_p99": 3.047723495797072e-8,
+        "abs_mean": 1.0266911358088638e-9,
+        "abs_p50": 1.4142287341201154e-10,
+        "abs_p95": 3.0405140627465244e-9,
+        "abs_p99": 2.25219967120438e-8,
         "rel_max": 5.177780320012974e-6,
-        "rel_mean": 4.3357289970588107e-10,
-        "rel_p50": 1.3931982528997815e-10,
-        "rel_p95": 1.3067644570630913e-9,
-        "rel_p99": 5.16707300247154e-9
+        "rel_mean": 3.9985892305619165e-10,
+        "rel_p50": 1.3934409121730228e-10,
+        "rel_p95": 1.298366831659975e-9,
+        "rel_p99": 4.015827932536157e-9
       },
       "samples_tested": 59003
     },
@@ -305,16 +305,16 @@
       "name": "asinh",
       "i16f16": {
         "count": 59007,
-        "abs_max": 0.009368954975967902,
-        "abs_mean": 0.0021266635260911025,
-        "abs_p50": 0.0014801997013025314,
-        "abs_p95": 0.006256595362091666,
-        "abs_p99": 0.008003853812445172,
-        "rel_max": 0.026957011464039424,
-        "rel_mean": 0.0006481167683768791,
-        "rel_p50": 0.0004840298276696915,
-        "rel_p95": 0.0017561901213065945,
-        "rel_p99": 0.002215571312015438
+        "abs_max": 0.009323178608780402,
+        "abs_mean": 0.0021284888817507635,
+        "abs_p50": 0.0014844659980624009,
+        "abs_p95": 0.006257894227960303,
+        "abs_p99": 0.007984492050545633,
+        "rel_max": 0.016260839471297594,
+        "rel_mean": 0.0006442542549104709,
+        "rel_p50": 0.0004833604162310917,
+        "rel_p95": 0.001753899668466157,
+        "rel_p99": 0.0022055481830100744
       },
       "i32f32": {
         "count": 59007,
@@ -335,16 +335,16 @@
       "name": "acosh",
       "i16f16": {
         "count": 59001,
-        "abs_max": 0.009733816681792629,
-        "abs_mean": 0.0022251709451035586,
-        "abs_p50": 0.0015877818855538628,
-        "abs_p95": 0.006443120819416226,
-        "abs_p99": 0.008309283221763941,
-        "rel_max": 0.002639144908644039,
-        "rel_mean": 0.0006738966410851339,
-        "rel_p50": 0.0005199648063320356,
-        "rel_p95": 0.0017993633344763151,
-        "rel_p99": 0.002270953061885943
+        "abs_max": 0.009688040314605129,
+        "abs_mean": 0.0022273528018676597,
+        "abs_p50": 0.0015914165377841627,
+        "abs_p95": 0.006447086883274444,
+        "abs_p99": 0.008295719356268272,
+        "rel_max": 0.002626733490764648,
+        "rel_mean": 0.0006736969672385671,
+        "rel_p50": 0.0005214367925851396,
+        "rel_p95": 0.0018006676087912823,
+        "rel_p99": 0.0022695168421162757
       },
       "i32f32": {
         "count": 59001,
@@ -365,16 +365,16 @@
       "name": "atanh",
       "i16f16": {
         "count": 59001,
-        "abs_max": 0.001042712906590637,
-        "abs_mean": 0.00005745413304488152,
-        "abs_p50": 0.000042381852155770616,
-        "abs_p95": 0.0001477479263376491,
-        "abs_p99": 0.0003605051872503928,
-        "rel_max": 0.6371890219738001,
-        "rel_mean": 0.0003884945546221465,
-        "rel_p50": 0.00007479710783610355,
-        "rel_p95": 0.0008262046645402671,
-        "rel_p99": 0.004189922942606896
+        "abs_max": 0.000996936539403137,
+        "abs_mean": 0.000048225767174518445,
+        "abs_p50": 0.00003234791334638665,
+        "abs_p95": 0.00013384945319838693,
+        "abs_p99": 0.00036007364172752077,
+        "rel_max": 0.3650807884541502,
+        "rel_mean": 0.0003009640696123398,
+        "rel_p50": 0.0000590390888308655,
+        "rel_p95": 0.0006253162143217303,
+        "rel_p99": 0.003043533591188758
       },
       "i32f32": {
         "count": 59001,
@@ -395,16 +395,16 @@
       "name": "acoth",
       "i16f16": {
         "count": 59001,
-        "abs_max": 0.0006742792358886973,
-        "abs_mean": 0.00004786209394856838,
-        "abs_p50": 0.00004119574939574766,
-        "abs_p95": 0.00011414091901112977,
-        "abs_p99": 0.00014234721267197652,
-        "rel_max": 0.014055554282461207,
-        "rel_mean": 0.0025039738676213056,
-        "rel_p50": 0.001578162208162982,
-        "rel_p95": 0.008079244902209818,
-        "rel_p99": 0.010661069956830615
+        "abs_max": 0.00063469181628939,
+        "abs_mean": 0.00003892003359818683,
+        "abs_p50": 0.00003349354136025079,
+        "abs_p95": 0.00009199005491542477,
+        "abs_p99": 0.00011657816953467709,
+        "rel_max": 0.01284496346263612,
+        "rel_mean": 0.0020995023707175223,
+        "rel_p50": 0.0013292638519312421,
+        "rel_p95": 0.006665326707366214,
+        "rel_p99": 0.009055939297598636
       },
       "i32f32": {
         "count": 59001,
@@ -426,26 +426,26 @@
       "i16f16": {
         "count": 59007,
         "abs_max": 0.46813349528929393,
-        "abs_mean": 0.008584803135554181,
-        "abs_p50": 0.000016375478616971473,
-        "abs_p95": 0.05339109960641508,
-        "abs_p99": 0.16808381906048453,
+        "abs_mean": 0.006142015618932389,
+        "abs_p50": 0.000014615890257533865,
+        "abs_p95": 0.03260455542874752,
+        "abs_p99": 0.1284817312330233,
         "rel_max": 0.3332355454934613,
-        "rel_mean": 0.011423961708018175,
-        "rel_p50": 0.00008313063134226579,
-        "rel_p95": 0.07867272148882169,
+        "rel_mean": 0.011417573122417932,
+        "rel_p50": 0.00006673575383040605,
+        "rel_p95": 0.07871431908682497,
         "rel_p99": 0.1835814167869461
       },
       "i32f32": {
         "count": 59007,
         "abs_max": 0.000014848937098577153,
-        "abs_mean": 4.591129582438186e-7,
-        "abs_p50": 3.602109027722733e-10,
-        "abs_p95": 3.300957587271114e-6,
-        "abs_p99": 7.671637376915896e-6,
+        "abs_mean": 4.428319596643952e-7,
+        "abs_p50": 2.8667312967911585e-10,
+        "abs_p95": 3.271608875365928e-6,
+        "abs_p99": 7.336121598200407e-6,
         "rel_max": 5.084182472060312e-6,
-        "rel_mean": 1.9143373009014e-7,
-        "rel_p50": 2.444515430731482e-9,
+        "rel_mean": 1.9135427833977108e-7,
+        "rel_p50": 2.2358414909904965e-9,
         "rel_p95": 1.2962977984451641e-6,
         "rel_p99": 3.2213312688224884e-6
       },
@@ -455,16 +455,16 @@
       "name": "ln",
       "i16f16": {
         "count": 59003,
-        "abs_max": 0.007028960622761815,
-        "abs_mean": 0.0000838118301668609,
-        "abs_p50": 0.00007251431725396884,
-        "abs_p95": 0.00020006211659051587,
-        "abs_p99": 0.0002573957974867369,
-        "rel_max": 0.08319029452329336,
-        "rel_mean": 0.00001986567678225745,
-        "rel_p50": 0.000012494171239045365,
-        "rel_p95": 0.00003892350594362073,
-        "rel_p99": 0.0000731862128693758
+        "abs_max": 0.007120513357136815,
+        "abs_mean": 0.00006108372519995839,
+        "abs_p50": 0.0000507042797313062,
+        "abs_p95": 0.0001528340420877683,
+        "abs_p99": 0.00020458018328906036,
+        "rel_max": 0.01244179019784243,
+        "rel_mean": 0.000013511077843052894,
+        "rel_p50": 8.759334514528923e-6,
+        "rel_p95": 0.00002967197208423552,
+        "rel_p99": 0.00005295682216701669
       },
       "i32f32": {
         "count": 59003,
@@ -485,16 +485,16 @@
       "name": "log2",
       "i16f16": {
         "count": 59003,
-        "abs_max": 0.0006140250690256366,
-        "abs_mean": 0.00011891466211653321,
-        "abs_p50": 0.00010307461606018364,
-        "abs_p95": 0.0002832318869696593,
-        "abs_p99": 0.0003662766333043521,
-        "rel_max": 0.011873289826984924,
-        "rel_mean": 0.00001771172892316616,
-        "rel_p50": 0.000012356254980832067,
-        "rel_p95": 0.000038052260832786956,
-        "rel_p99": 0.00007053468077128446
+        "abs_max": 0.0006598014362131366,
+        "abs_mean": 0.00008558332538114927,
+        "abs_p50": 0.00007083049510292483,
+        "abs_p95": 0.00021631260077370484,
+        "abs_p99": 0.00028929750223483097,
+        "rel_max": 0.014636494762760797,
+        "rel_mean": 0.000013259536025697493,
+        "rel_p50": 8.478787777090254e-6,
+        "rel_p95": 0.000029151110902145505,
+        "rel_p99": 0.00005287799348650687
       },
       "i32f32": {
         "count": 59003,
@@ -515,16 +515,16 @@
       "name": "log10",
       "i16f16": {
         "count": 59003,
-        "abs_max": 0.00018310546875,
-        "abs_mean": 0.000037289963196447204,
-        "abs_p50": 0.00003219926275743745,
-        "abs_p95": 0.00008958987203810942,
-        "abs_p99": 0.00011631138507617322,
-        "rel_max": 0.013356845368307679,
-        "rel_mean": 0.0000186368528692315,
-        "rel_p50": 0.000012807206636373906,
-        "rel_p95": 0.00003986763539042081,
-        "rel_p99": 0.00007370750128702391
+        "abs_max": 0.0001983642578125,
+        "abs_mean": 0.00002802985157528529,
+        "abs_p50": 0.000023431607028001622,
+        "abs_p95": 0.00007010842908883319,
+        "abs_p99": 0.000092947326506998,
+        "rel_max": 0.020974729725972013,
+        "rel_mean": 0.000014423993800471001,
+        "rel_p50": 9.28207260892599e-6,
+        "rel_p95": 0.000031374333977432835,
+        "rel_p99": 0.00005676337528031852
       },
       "i32f32": {
         "count": 59003,
@@ -546,28 +546,28 @@
       "i16f16": {
         "count": 59007,
         "abs_max": 0.1240198076766319,
-        "abs_mean": 0.0034303713742522895,
-        "abs_p50": 0.000039326024889874134,
-        "abs_p95": 0.02136241704067743,
-        "abs_p99": 0.054646100538207065,
+        "abs_mean": 0.003254876418514731,
+        "abs_p50": 0.000029241600689067226,
+        "abs_p95": 0.019762391636334087,
+        "abs_p99": 0.052000798585595476,
         "rel_max": 0.015141079259996712,
-        "rel_mean": 0.0007346566450210694,
-        "rel_p50": 0.00006021458901555424,
-        "rel_p95": 0.004702059500442153,
+        "rel_mean": 0.0007282995676444927,
+        "rel_p50": 0.00004740369660883063,
+        "rel_p95": 0.004705167104145861,
         "rel_p99": 0.01027006938792182
       },
       "i32f32": {
         "count": 59007,
         "abs_max": 2.711145384637348e-6,
-        "abs_mean": 7.240070011031757e-8,
-        "abs_p50": 8.620442137896589e-10,
-        "abs_p95": 4.68344865112158e-7,
-        "abs_p99": 1.1136039574921597e-6,
-        "rel_max": 4.44557716494002e-6,
-        "rel_mean": 1.1722305125266224e-8,
-        "rel_p50": 1.2452105978093034e-9,
-        "rel_p95": 7.3844316765736e-8,
-        "rel_p99": 1.569632910531003e-7
+        "abs_mean": 6.906259627837998e-8,
+        "abs_p50": 6.68742838882963e-10,
+        "abs_p95": 4.3179491626688105e-7,
+        "abs_p99": 1.0690903309296118e-6,
+        "rel_max": 2.384185791015625e-7,
+        "rel_mean": 1.1514619376993808e-8,
+        "rel_p50": 1.0561495138250126e-9,
+        "rel_p95": 7.384199695898233e-8,
+        "rel_p99": 1.5696090047105027e-7
       },
       "samples_tested": 59007
     },