INDEX

Explanations

it was

The neuron fires on short existential or perceptual “discovery” phrases—e.g. “there it is,” “I see him,” or “I can feel it in my bones.”

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 travailleurs

-1.44

verture

-1.31

 représentants

-1.30

 撮っ

-1.30

joey

-1.27

ureusement

-1.27

 Mientras

-1.26

ventus

-1.24

いています

-1.24

-1.23

POSITIVE LOGITS

ється

1.18

It

1.17

()]);

1.14

For

1.13

 like

1.10

At

1.09

 gave

1.09

any

1.09

);

1.09

An

1.07

Activations Density 0.049%