INDEX

Explanations

counting studies or reports

The neuron activates on numeric quantitative mentions—digits and the count nouns or measurement terms they accompany.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

or

-1.77

at

-1.67

 without

-1.61

to

-1.59

and

-1.49

 where

-1.41

no

-1.40

 just

-1.40

 very

-1.39

an

-1.35

POSITIVE LOGITS

 refugiados

1.46

 mélanger

1.38

 både

1.38

 tačiau

1.37

 fidé

1.35

 appeler

1.30

 relâche

1.28

 méda

1.26

 they

1.25

 verrou

1.24

Activations Density 0.230%