INDEX

Explanations

call me

This neuron detects the phrase “call me” (i.e. direct first‐person invitations to be contacted).

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

 Whether

-1.87

Its

-1.81

 When

-1.79

я

-1.73

 عليك

-1.71

 Asking

-1.70

 That

-1.65

you

-1.65

ernalia

-1.62

 Those

-1.61

POSITIVE LOGITS

2.45

馓

1.89

 rappeler

1.88

</h1>

1.81

!’

1.80

</b>

1.75

 станет

1.73

,’

1.68

 êtres

1.66

 trecho

1.65

Activations Density 0.010%