TSR generates error

[hugo-rddi]

Hello, Some errors are occurring during TSR execution. We tried different configurations (tried the different saliency methods available, FO, IG, SVS, ...).

Context:

The following cell worked.

int_mod = TSR(
    model, 
    train_x.shape[-1],
    train_x.shape[-2], 
    method='IG',
    mode='feat'
)

But this one didn't :

exp=int_mod.explain(
    input_x,
    labels=class_target,
    TSR=True
)

Error :

    147 elif self.method == "IG":
    148     base = baseline_single
--> 149     attributions = self.Grad.attribute(
    150         input, baselines=baseline_single, target=labels
    151     )
    152 elif self.method == "DL":
    153     base = baseline_single

     41 def wrapper(*args, **kwargs):
---> 42     return func(*args, **kwargs)
...
    527         )
    528 else:
--> 529     raise AssertionError(f"Target type {type(target)} is not valid.")

AssertionError: Target type  is not valid.

Any idea what can be the problem here ? Model, train_x, labels and so on are not the origin of the problem given the fact that they respect the same format as expected and worked in NativeGuide execution.

Thanks in advance !

[JHoelli]

Hi,

did you parse the label explicitly into int ? int(np.argmax(y_target, axis=1)[0])`

The base library (captum) is unfortunately pretty strict with that. Really appreciate your feedback, it really helps us to improve the library. I will add some preprocessing for that in the next release. So that others do not run into the same issues.

[JHoelli]

I added some preprocessing that should solve it: pip install https://github.com/fzi-forschungszentrum-informatik/TSInterpret/archive/refs/heads/main.zip

As soon as I have additional time I will also improve the error handeling (I know there is still a lot to do ;)).

[hugo-rddi]

Again you were right ! Needed to be strictly casted as int. Thanks, have a nice day

[JHoelli]

Nice.