INDEX

Explanations

ideas, websites, myths, food

The neuron fires on technical or code‐related terms in software documentation—words like “software,” “integrates,” “platform,” “UI,” and similar development jargon.

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

as

-1.77

 along

-1.57

and

-1.57

 together

-1.45

along

-1.34

together

-1.22

 ainsi

-1.17

 particularly

-1.15

 while

-1.09

called

-1.07

POSITIVE LOGITS

 przypadku

1.43

 before

1.25

Before

1.24

 there

1.22

 whose

1.20

Opere

1.19

每个

1.17

 경우

1.15

そうですが

1.14

There

1.13

Activations Density 0.037%