INDEX

Explanations

finally summing up

The neuron strongly activates on numerical values and measurements (digits, stats, quantities).

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

并

0.80

并在

0.79

and

0.79

与其

0.75

及び

0.75

とその

0.75

并将

0.73

 及び

0.71

and

0.69

 এবং

0.67

POSITIVE LOGITS

等等

1.58

 многое

1.50

 sebagainya

1.49

 sogar

1.39

etc

1.37

 whatnot

1.37

 ultimately

1.35

 much

1.34

 даже

1.34

 навіть

1.34

Activations Density 0.602%