INDEX

Explanations

reading specific texts

The neuron strongly responds to mentions of reading (especially the verb “read” in contexts like “I read this/story/blog…”).

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

 instead

-1.38

 rather

-1.21

幛

-1.20

selben

-1.16

 adopción

-1.15

};

-1.13

 plutôt

-1.12

ны

-1.10

ˋ

-1.10

 гостиной

-1.07

POSITIVE LOGITS

did

1.64

 came

1.47

依照

1.43

 were

1.35

 began

1.29

was

1.24

 went

1.23

 yesterday

1.21

 originally

1.20

 took

1.20

Activations Density 0.111%