INDEX

Explanations

from what i

This neuron detects the “From what I …” reporting phrase (e.g. “From what I can tell,” “From what I gathered,” etc.).

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

");

-1.73

 chapéu

-1.66

 campeão

-1.58

 Likely

-1.57

 héroe

-1.55

 relatório

-1.53

 ejército

-1.52

-1.47

 asesino

-1.47

 inolvid

-1.46

POSITIVE LOGITS

of

2.16

or

1.98

hline

1.73

 them

1.71

벤트

1.70

 utilizzati

1.63

 blauw

1.61

it

1.60

vání

1.59

 FirstName

1.59

Activations Density 0.006%