jax-ml/jax · error · TypeError
val.dtype= != ref.dtype=
Error message
val.dtype={x.dtype} != ref.dtype={ref.dtype} What it means
Validation in the SparseCore scatter abstract eval: the stored value x must have exactly the ref's dtype (int32 or float32 refs only); a dtype mismatch between the value and the ref raises TypeError. The input at fault is the x argument of the scatter.
Solutions
- Cast x to ref.dtype before the scatter (x.astype(ref_dtype))
- Or declare the ref with the same dtype as your computed values
Example fix
// before sc_primitives.store_scatter(ref, indices, x) # x f32, ref int32 // after sc_primitives.store_scatter(ref, indices, x.astype(ref.dtype))
Defensive patterns
Strategy: type-guard
Validate before calling
if x.dtype != ref.dtype:
x = x.astype(ref.dtype) Type guard
def dtypes_match(x, ref) -> bool:
return x.dtype == getattr(getattr(ref, 'aval', ref), 'dtype', None) Prevention
- Always cast x to ref.dtype right before scatter
- Avoid mixing compute dtype and buffer dtype in SC kernels
- Add asserts in kernel tests covering dtype agreement
When it happens
Trigger: x is e.g. float32 while the ref was declared int32 (or vice versa), or bf16 value into f32 ref.
Common situations: Mixing computation dtype (often bf16/f32) with a differently-typed output buffer; forgetting that Pallas requires exact dtype agreement.
Related errors
- Indices must not be empty
- does not match expected shape
- Scatter only supports storing to VMEM, got
- Scatter only supports VectorSubcoreMesh, got
- does not match expected shape
AI-assisted analysis of jax-ml/jax@1e1c6a8fc0 (2026-08-27).
Data as JSON: /api/errors/050affc7e75aae09.
Report an issue: GitHub.
Appendix: source
Thrown at jax/_src/pallas/mosaic/sc_primitives.py:332
scatter_p = jax_core.Primitive("scatter")
scatter_p.is_effectful = lambda params: True
scatter_p.multiple_results = True
@scatter_p.def_effectful_abstract_eval
def _scatter_abstract_eval(*flat_args, tree, add):
ref, transforms, indices, x, mask = jax.tree.unflatten(tree, flat_args)
if transforms:
ref = state_types.TransformedRef(ref, transforms)
if ref.dtype not in (jnp.int32, jnp.float32):
raise TypeError(f"ref.dtype={ref.dtype} must be int32 or float32")
expected_shape = _indexed_shape(ref, indices)
if x.shape != expected_shape:
raise ValueError(
f"{x.shape=} does not match expected shape {expected_shape}"
)
if x.dtype != ref.dtype:
raise TypeError(f"val.dtype={x.dtype} != ref.dtype={ref.dtype}")
if mask is not None:
if mask.shape != expected_shape:
raise ValueError(
f"{mask.shape=} does not match expected shape {expected_shape}"
)
if mask.dtype != jnp.bool:
raise TypeError(f"Mask must be a boolean array, got {mask.dtype}")
effects: set[jax_core.Effect] = {state_types.WriteEffect(0)}
if add:
effects.add(state_types.ReadEffect(0))
return (), effects
@sc_lowering.register_lowering_rule(scatter_p)
def _scatter_lowering_rule(
ctx: sc_lowering.LoweringRuleContext, *flat_args, tree, add
):
ref, transforms, indices, x, mask = jax.tree.unflatten(tree, flat_args)View on GitHub (pinned to 1e1c6a8fc0)