keras-team/keras · error · ValueError

Argument `sensitivity` must be in the range [0, 1]. Received

Error message

Argument `sensitivity` must be in the range [0, 1]. Received: sensitivity={sensitivity}

What it means

Raised in the constructor of keras.metrics.SensitivityAtSpecificity when the sensitivity argument is below 0 or above 1. Sensitivity (recall / true positive rate) is a probability, so Keras validates the range eagerly in __init__.

Source

Thrown at keras/src/metrics/confusion_metrics.py:847

    ```python
    model.compile(
        optimizer='sgd',
        loss='binary_crossentropy',
        metrics=[keras.metrics.SpecificityAtSensitivity(sensitivity=0.3)])
    ```
    """

    def __init__(
        self,
        sensitivity,
        num_thresholds=200,
        class_id=None,
        name=None,
        dtype=None,
    ):
        if sensitivity < 0 or sensitivity > 1:
            raise ValueError(
                "Argument `sensitivity` must be in the range [0, 1]. "
                f"Received: sensitivity={sensitivity}"
            )
        self.sensitivity = sensitivity
        self.num_thresholds = num_thresholds
        super().__init__(
            sensitivity,
            num_thresholds=num_thresholds,
            class_id=class_id,
            name=name,
            dtype=dtype,
        )

    def result(self):
        sensitivities = ops.divide_no_nan(
            self.true_positives,
            ops.add(self.true_positives, self.false_negatives),
        )

View on GitHub (pinned to 7a34a03db6)

Solutions

1. Use a fraction in [0, 1], e.g. sensitivity=0.95.
2. Validate config values before metric construction.
3. Ensure upstream calculations produce probabilities, not percentages.

Example fix

# before
m = keras.metrics.SensitivityAtSpecificity(specificity=0.5, sensitivity=95)

# after
m = keras.metrics.SensitivityAtSpecificity(specificity=0.5, sensitivity=0.95)

if not (0.0 <= sensitivity <= 1.0):
    raise ValueError(f'sensitivity must be in [0,1], got {sensitivity}')

def is_probability(v) -> bool:
    return isinstance(v, (int, float)) and 0.0 <= v <= 1.0

When it happens

Trigger: keras.metrics.SensitivityAtSpecificity(sensitivity=1.5); passing 95 (percent) instead of 0.95; values from unnormalized computations.

Common situations: Percent/fraction mix-ups; ported code where targets were percentages; config typos.

Related errors

• Argument `specificity` must be in the range [0, 1]. Received
• Argument `recall` must be in the range [0, 1]. Received: rec
• Argument `precision` must be in the range [0, 1]. Received:
• Invalid `beta` argument value. It should be > 0. Received: b
• Invalid `threshold` argument value. It should verify 0 < thr

AI-assisted analysis of keras-team/keras@7a34a03db6 (2026-08-25). Data as JSON: /api/errors/413feef2325e050b. Report an issue: GitHub.