INDEX

Explanations

This neuron strongly activates on end-of-sentence punctuation marks (e.g. periods, exclamation points) rather than words.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Asimismo

0.38

Furthermore

0.38

Moreover

0.37

cknowledg

0.33

 beispielsweise

0.33

Referências

0.33

0.32

References

0.32

Mutability

0.32

Instances

0.32

POSITIVE LOGITS

 Didn

0.74

 Seems

0.71

 VERY

0.69

 didn

0.67

 Очень

0.66

 took

0.64

 Looks

0.64

 Doesn

0.64

 Took

0.63

 очень

0.62

Activations Density 1.874%