INDEX

Explanations

The neuron fires on tokens that begin new clauses or sentences—that is, on words (or numbers) immediately following punctuation or section breaks.

made up of

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

presumably

-1.27

itaji

-1.27

resaid

-1.24

 проє

-1.21

://"

-1.20

 superbes

-1.20

clearly

-1.13



-1.13

 relevante

-1.13

 Recomend

-1.11

POSITIVE LOGITS

and

1.55

 when

1.27

 both

1.24

 referirse

1.16

ulary

1.16

 this

1.15

 water

1.12

諄

1.12

Both

1.12

This

1.09

Activations Density 0.045%