Edit implementation of jax.numpy.ldexp to get correct gradient.

CHANGELOG.md

@@ -60,6 +60,7 @@ When releasing, please add the new-release-boilerplate to docs/pallas/CHANGELOG.
 * Bug fixes
   * Fixed a bug where {func}`jax.numpy.cumsum` would produce incorrect outputs
     if a non-boolean input was provided and `dtype=bool` was specified.
+  * Edit implementation of {func}`jax.numpy.ldexp` to get correct gradient.
 
 ## jax 0.4.33 (September 16, 2024)
 

jax/_src/numpy/ufuncs.py

@@ -2451,34 +2451,7 @@ def ldexp(x1: ArrayLike, x2: ArrayLike, /) -> Array:
   x1 = lax.convert_element_type(x1, dtype)
   x2 = lax.convert_element_type(x2, int_type)
 
-  mask = (1 << info.nexp) - 1
-  bias = 1 - info.minexp
-  x, e = _normalize_float(x1)
-  x2 += e + ((x >> info.nmant) & mask) - bias
-
-  # find underflow/overflow before denormalization
-  underflow_cond = less(x2, -(bias + info.nmant))
-  overflow_cond = greater(x2, bias)
-
-  m = lax.full_like(x, 1, dtype=dtype)
-
-  # denormals
-  cond = less(x2, -bias + 1)
-  x2 = _where(cond, x2 + info.nmant, x2)
-  m = _where(cond, m / (1 << info.nmant), m)
-
-  x2 = lax.convert_element_type(x2, np.int32)
-  x &= ~(mask << info.nmant)
-  x |= ((lax.convert_element_type(x2, int_type) + bias) << info.nmant)
-
-  x = lax.convert_element_type(m, dtype) * lax.bitcast_convert_type(x, dtype)
-
-  # underflow
-  x = _where(underflow_cond, lax.full_like(x, 0, dtype=dtype), x)
-  # overflow
-  x = _where(overflow_cond, lax.sign(x1) * lax.full_like(x, np.inf), x)
-  # ldexp(x1, x2) = x1 for x1 = inf, -inf, nan, 0
-  return _where(isinf(x1) | isnan(x1) | (x1 == 0), x1, x)
+  return x1 * lax.convert_element_type(2 ** x2, dtype)
 
 
 @implements(np.frexp, module='numpy')

tests/lax_numpy_test.py

@@ -6142,6 +6142,15 @@ def testGradLogaddexp2Complex(self, shapes, dtype):
       tol = 3e-2
     check_grads(jnp.logaddexp2, args, 1, ["fwd", "rev"], tol, tol)
 
+  @jtu.sample_product(
+    n=range(-4, 5),
+    dtype=[jnp.float32, jnp.float64],
+  )
+  def testGradLdexp(self, n, dtype):
+    rng = jtu.rand_default(self.rng())
+    x = rng((), dtype)
+    check_grads(lambda x: jnp.ldexp(x, n), (x,), 1)
+
 
 class NumpySignaturesTest(jtu.JaxTestCase):