INDEX

Explanations

unique properties

The neuron responds strongly to declarative, copular “X is/are Y” statements that define or describe a subject.

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

unno

-1.48

vermectin

-1.14

ebly

-1.13

UPAC

-1.09

ridescent

-1.07

strual

-1.05

ggy

-1.04

carus

-1.02

 instead

-1.01

接続

-0.98

POSITIVE LOGITS

 requires

1.27

Requires

1.16

 need

1.09

 membutuhkan

1.07

 Requires

1.07

独特

1.04

abaikan

1.03

Ә

1.02

 require

1.02

 atve

1.01

Activations Density 0.050%