INDEX

Explanations

Japan, Main, wood

The neuron predominantly activates on open‐class content words—especially common nouns—in running prose.

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

 бампер

-1.41

 πρ

-1.30

ngulo

-1.29

när

-1.22

 goû

-1.20

ﻷ

-1.19

 szab

-1.17

 TextFormField

-1.16

butuhkan

-1.14

 заменить

-1.14

POSITIVE LOGITS

1.41

 from

1.35

1.32

 before

1.31

［

1.25

 after

1.22

[]

1.21

 they

1.18

he

1.17

.°

1.16

Activations Density 0.014%