INDEX

Explanations

broad-based, depth behind, physical brand

The neuron is detecting tokens bearing common derivational or inflectional suffixes (e.g. –s, –ed, –ness, –ies, –ical, –based).

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

and

-2.16

of

-1.95

if

-1.91

by

-1.89

-1.86

 Behaup

-1.85

 based

-1.82

 sicheren

-1.80

 after

-1.75

as

-1.71

POSITIVE LOGITS

鶿

2.27

Estos

2.06

疴

1.97

voer

1.85

𖧵

1.82

didSet

1.80

 welkom

1.79

ኧ

1.75

Aquí

1.74

 appris

1.74

Activations Density 0.042%