{% for group, issues in issues_by_group.items() %}
We found some data slices in your dataset on which your model performance is lower than average. Performance bias may happen for different reasons:
To learn more about causes and solutions, check our guide on performance bias.
{% elif issues[0].__class__.__name__ == "RobustnessIssue" %}Your model seems to be sensitive to small perturbations in the input data. These perturbations can include adding typos, changing word order, or turning text into uppercase or lowercase. This happens when:
To learn more about causes and solutions, check our guide on robustness issues.
{% elif issues[0].__class__.__name__ == "OverconfidenceIssue" %}We found some data slices in your dataset containing significant number of overconfident predictions. Overconfident predictions are rows that are incorrect but are predicted with high probabilities or confidence scores. This happens when:
To learn more about causes and solutions, check our guide on overconfidence issues.
{% elif issues[0].__class__.__name__ == "UnderconfidenceIssue" %}We found some data slices in your dataset containing significant number of underconfident predictions. Underconfident predictions refer to situations where the predicted label has a probability that is very close to the probability of the next highest probability label. This happens when:
To learn more about causes and solutions, check our guide on underconfidence issues.
{% elif issues[0].__class__.__name__ == "EthicalIssue" %}Your model seems to be sensitive to gender, ethnic, or religion based perturbations in the input data. These perturbations can include switching some words from feminine to masculine, countries or nationalities. This happens when:
To learn more about causes and solutions, check our guide on unethical behaviour.
{% elif issues[0].__class__.__name__ == "DataLeakageIssue" %}Your model seems to present some data leakage. The model provides different results depending on whether it is computing on a single data point or the entire dataset. This happens when:
To learn more about causes and solutions, check our guide on data leakage.
{% elif issues[0].__class__.__name__ == "StochasticityIssue" %}Your model seems to present some stochastic behaviour. The model provides different results at each execution. This may happen when some stochastic training process is included in the prediction pipeline.
To learn more about causes and solutions, check our guide on stochasticity issues.
{% elif issues[0].__class__.__name__ == "LLMToxicityIssue" %}Your model seems to exhibit offensive behaviour when we use adversarial “Do Anything Now” (DAN) prompts.
{% elif issues[0].__class__.__name__ == "HarmfulnessIssue" %}Your model can generate harmful or toxic content in response to specific prompts.
{% elif issues[0].__class__.__name__.endswith("StereotypeIssue") %}Your model exhibits social stereotypes about genders or minorities.
{% elif issues[0].__class__.__name__ == "SpuriousCorrelationIssue" %}We found some potential spurious correlations between your data and the model predictions. Some data slices are highly correlated with your predictions. This happens when:
To learn more about causes and solutions, check our guide on spurious correlation.
{% else %}Found issues for {{ issues[0].group }}
{% endif %}We found no issues in your model. Good job!