INDEX

Explanations

by the time

The neuron strongly activates on personal pronouns (and occasionally the word “time”), i.e. tokens that refer directly to people (you, he, they, them).

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

-2.45

$\

-2.05

and

-1.87

 mecánica

-1.78

ла

-1.77

-1.76

");

-1.75

}(

-1.72

 комнаты

-1.71

$>$

-1.70

POSITIVE LOGITS

you

2.09

poň

2.03

某个

1.92

 diinginkan

1.83

苇

1.81

ل

1.80

した後

1.78

した場合

1.77

哂

1.75

 selepas

1.74

Activations Density 0.002%