INDEX

Explanations

rejects or dismisses

The neuron fires on words that denote formal rejection or dismissal (e.g. “dismiss,” “rejected,” “dismissed”) in a legal context.

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

to

-1.11

 να

-0.90

ゆっくり

-0.89

 mentre

-0.88

verk

-0.87

 freiw

-0.85

oa

-0.84

 Worte

-0.83

 consenti

-0.79

ktiv

-0.76

POSITIVE LOGITS

out

1.87

 summar

1.55

 outright

1.49

 uncer

1.20

 unread

1.09

 without

1.07

as

1.05

 cavalier

0.98

 wholesale

0.97

Chas

0.92

Activations Density 0.028%