Can't get the gradient attribution map to work from "Explore Network Predictions Using Deep Learning Visualization Techniques" page on Mathworks

I am using a custom trained AlexNet which has a softmaxname called "prob" but get the following error message:

Error using nnet.internal.cnn.dlnetwork>iAssertLayerNameExist

Layer 'prob' does not exist.

Error in nnet.internal.cnn.dlnetwork>iGetOutputIndices (line 497)

iAssertLayerNameExist(layerName, map);

Error in nnet.internal.cnn.dlnetwork/getOutputLayersIndicesFromNames (line 292)

[outputLayerIdx(i), outputLayerPortIdx(i)] = iGetOutputIndices(outputNames(i), layerNameMap, layers);

Error in dlnetwork/getOutputLayersIndices (line 1228)

[layerIndices, layerOutputIndices] = getOutputLayersIndicesFromNames(net.PrivateNetwork, string(outputNames));

Error in dlnetwork/predict (line 859)

[layerIndices, layerOutputIndices] = getOutputLayersIndices(net, NameValueArgs);

Error in gradientMap (line 7)

scores = predict(dlnet,I,'Outputs',{softmaxName});

Error in deep.internal.dlfeval (line 17)

[varargout{1:nargout}] = fun(x{:});

Error in dlfeval (line 40)

[varargout{1:nargout}] = deep.internal.dlfeval(fun,varargin{:});

Error in gradientAttribution (line 15)

dydI = dlfeval(@gradientMap,dlnet,dlImg,softmaxName,YPred);

Error in untitled14 (line 48)

pixelMap = gradientAttribution(netTransfergrey10dB,img,YPred,softmaxName,"autodiff");

[Varun]

Hi, can you please share your implementation of the AlexNet?