keras-team/keras · error · ValueError

Unsupported sparse format: {x1.__class__}

Error message

Unsupported sparse format: {x1.__class__}

What it means

Error "Unsupported sparse format: {x1.__class__}" thrown in keras-team/keras.

Source

Thrown at keras/src/backend/jax/sparse.py:322

                        x2_zeros_and_nans = jnp.equal(x2, 0)
                        if not jnp.issubdtype(x2.dtype, jnp.integer):
                            x2_zeros_and_nans = jnp.logical_or(
                                x2_zeros_and_nans, jnp.isnan(x2)
                            )
                        # 2. Make it a BCOO of True values.
                        x2_zeros_and_nans = jax_sparse.bcoo_fromdense(
                            x2_zeros_and_nans,
                            n_batch=x1.n_batch,
                            n_dense=x1.n_dense,
                            index_dtype=x1.indices.dtype,
                        )
                        # 3. Add the indices to x1.
                        x1 = bcoo_add_indices(
                            x1, x2_zeros_and_nans, sum_duplicates=True
                        )
                    return sparse_func(x1, x2)
            else:
                raise ValueError(f"Unsupported sparse format: {x1.__class__}")
        elif isinstance(x2, jax_sparse.JAXSparse):
            # x1 is dense, x2 is sparse, densify x2
            x2 = x2.todense()
        return func(x1, x2)

    return sparse_wrapper

View on GitHub (pinned to 7a34a03db6)

When it happens

Trigger: Thrown at keras/src/backend/jax/sparse.py:322 when the library encounters an invalid state.

Common situations: See trigger scenarios.

AI-assisted analysis of keras-team/keras@7a34a03db6 (2026-08-25). Data as JSON: /api/errors/403e21e518341029. Report an issue: GitHub.