INDEX

Explanations

actual followed by noun

The neuron fires on the use of the intensifying modifier “actual,” marking words emphasized for authenticity.

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

it

-2.28

 where

-1.84

</td>

-1.73

is

-1.73

</b>

-1.73

nargs

-1.69

each

-1.63

</h5>

-1.61

 ไม่

-1.61

-1.60

POSITIVE LOGITS

 schwach

1.77

推荐阅读

1.71

本日の

1.69

?”

1.68

图案

1.67

鐫

1.63

-“

1.63

富有

1.60

,“

1.59

墙壁

1.55

Activations Density 0.014%