Analyze model confidence for JSONL files with question/answer data
Processing data...
The model predicted this token with a probability of 0.00.
Higher probability values indicate greater confidence in the predicted token.
This token has an entropy value of 0.00.
Higher entropy values indicate greater uncertainty in the prediction.
This token has an Aleatoric Uncertainty (AU) value of 0.00.
AU represents the inherent noise in the data that cannot be reduced.
This token has an Epistemic Uncertainty (EU) value of 0.00.
EU represents uncertainty due to limited knowledge and can be reduced with more data.
These are the other most likely tokens the model considered (logits):
Rank | Token ID | Logit Value |
---|
Full JSON data for the selected question:
This website has been generated by DeepSite