INDEX

Explanations

introducing examples

The neuron spotlights discourse-structuring words and phrases (e.g., “First,” “Then,” “Because,” “In fact,” “Unfortunately,” “Example,” etc.) that signal transitions or logical connections in text.

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

 pantalón

-1.29

__":

-1.25

(!__

-1.22

and

-1.13

燈

-1.05

*$

-1.04

的需求

-1.02

beeld

-1.02

asily

-1.02

glected

-1.02

POSITIVE LOGITS

1.30

 including

1.27

 even

1.26

 their

1.12

 because

1.09

 this

1.08

the

1.08

in

1.04

 since

1.00

Psychiat

0.98

Activations Density 0.003%