INDEX

Explanations

innocent people

The neuron fires on the main content words (nouns, verbs, adjectives) of the book‐review passages.

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

 poë

-1.43

わざわざ

-1.30

心头

-1.29

亵

-1.23

</em>

-1.23

参戦

-1.20

ּוֹ

-1.19

的说道

-1.19

咂

-1.16

</strong>

-1.16

POSITIVE LOGITS

虽然

1.39

endet

1.31

 景色

1.24

雖然

1.23

随着

1.23

 their

1.16

 vantaggi

1.16



1.15

 είχε

1.13

she

1.13

Activations Density 0.027%