INDEX

Explanations

The neuron detects verbs that indicate modifying or working on content (e.g. “edit,” “work on,” “alter”).

pronouns followed by conjunctions

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

of

-1.13

 doesnt

-1.09

no

-1.09

man

-1.02

-1.01

ఱ

-1.01

大丈夫です

-1.00

ab

-1.00

 suyu

-0.98

setAll

-0.98

POSITIVE LOGITS

 into

2.08

 with

1.28

by

1.23

というわけで

1.20

 modele

1.13

 through

1.13

}}$.

1.12

ではまた

1.12

dekat

1.10

 大全

1.07

Activations Density 0.056%