INDEX

Explanations

prerequisites and requirements

This neuron detects expressions of necessity or requirement, such as “needed,” “required,” or similar obligation verbs introducing what must be done.

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

卨

-1.73

 have

-1.60

廰

-1.53

Produzione

-1.50

臯

-1.47

Hintergrund

-1.38

 levante

-1.38

Weiterlesen

-1.38

をします

-1.37

 remplacement

-1.33

POSITIVE LOGITS

to

2.20

in

1.84

for

1.77

but

1.59

 republiky

1.52

↵↵

1.46

 Neither

1.45

 when

1.44

事的

1.43

":[{

1.41

Activations Density 0.039%