INDEX

Explanations

word or phrase connector

The neuron fires on function words and markers typical of dictionary or etymology entries (e.g. like, meaning, Latin, Greek, common, commas, and).

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

隳

-1.27

 renon

-1.25

 beper

-1.21

 totalement

-1.20

 suffisamment

-1.20

 rése

-1.20

おわりに

-1.20

܀

-1.19

 pratiquement

-1.19

 ſever

-1.18

POSITIVE LOGITS

 after

1.45

 although

1.38

 with

1.30

 from

1.27

 while

1.27

in

1.26

any

1.23

1.21

1.16

at

1.10

Activations Density 0.023%