Is the following problem an issue with Google Colab or an error in the code?

I am trying to follow "Deep Learning in Python" by Francois Chollet.

I have copied and pasted the code at the end of this post from the github page into colab. I am getting a very different, and in my opinion, incorrect image. I should be obtaining:

when in fact I am obtaining:

Here is the code, copied from https://github.com/fchollet/deep-learning-with-python-notebooks/blob/master/chapter10_interpreting-what-convnets-learn.ipynb

Is this a problem with Colab?

img_path = keras.utils.get_file(
    fname="elephant.jpg",
    origin="https://img-datasets.s3.amazonaws.com/elephant.jpg",
)
img = keras.utils.load_img(img_path)
img_array = np.expand_dims(img, axis=0)

model = keras_hub.models.ImageClassifier.from_preset(
   "xception_41_imagenet",
   activation="softmax",
)
preds = model.predict(img_array)
preds.shape

keras_hub.utils.decode_imagenet_predictions(preds)

last_conv_layer_name = "block14_sepconv2_act"
last_conv_layer = model.backbone.get_layer(last_conv_layer_name)
last_conv_layer_model = keras.Model(model.inputs, last_conv_layer.output)

classifier_input = last_conv_layer.output
x = classifier_input
for layer_name in ["pooler", "predictions"]:
    x = model.get_layer(layer_name)(x)
classifier_model = keras.Model(classifier_input, x)

import tensorflow as tf

def get_top_class_gradients(img_array):
    last_conv_layer_output = last_conv_layer_model(img_array)
    with tf.GradientTape() as tape:
        tape.watch(last_conv_layer_output)
        preds = classifier_model(last_conv_layer_output)
        top_pred_index = ops.argmax(preds[0])
        top_class_channel = preds[:, top_pred_index]

    grads = tape.gradient(top_class_channel, last_conv_layer_output)
    return grads, last_conv_layer_output

grads, last_conv_layer_output = get_top_class_gradients(img_array)
grads = ops.convert_to_numpy(grads)
last_conv_layer_output = ops.convert_to_numpy(last_conv_layer_output)


pooled_grads = np.mean(grads, axis=(0, 1, 2))
last_conv_layer_output = last_conv_layer_output[0].copy()
for i in range(pooled_grads.shape[-1]):
    last_conv_layer_output[:, :, i] *= pooled_grads[i]
heatmap = np.mean(last_conv_layer_output, axis=-1)

heatmap = np.maximum(heatmap, 0)
heatmap /= np.max(heatmap)
plt.matshow(heatmap)