INDEX

Explanations

legal, credit, or truck contexts

The neuron fires strongly on verbs and nouns describing formal legal or bureaucratic procedural steps (e.g. filed, claims, benefits, examiner, After, Following, Defendant was indicted).

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

 already

-1.27

 such

-1.21

if

-1.20

 another

-1.16

 will

-1.12

zí

-1.10

one

-1.09

already

-1.05

may

-1.03

Already

-1.00

POSITIVE LOGITS

 then

1.66

 سپس

1.23

ִּ

1.23

ַּ

1.21

เมื่อ

1.13

 subsequently

1.11

 thereupon

1.09

ֶּ

1.05

ブログ

1.03

then

1.02

Activations Density 0.076%