INDEX

Explanations

descriptive terms

The neuron strongly activates on descriptive adjectives—especially those ending in “‐ant,” “‐ent,” or “‐ive”—that qualify or characterize nouns.

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

The

-2.53

 แต่

-2.20

是啊

-1.90

黑衣

-1.70

 WHEN

-1.69

好人

-1.66

 ถ้า

-1.64

kkor

-1.63

 Würde

-1.63

 Quando

-1.61

POSITIVE LOGITS

2.41

that

2.23

is

2.11

块钱

2.05

(?)

1.95

 kwamba

1.93

1.92

and

1.90

—

1.86

 sebagainya

1.83

Activations Density 0.110%