INDEX

Explanations

modern or current concepts

The neuron fires on adjectives and modifiers that signal prominence, novelty, or historical significance (e.g. “modern,” “current,” “first,” “major,” “best,” etc.).

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

or

-1.51

 dlaczego

-1.41

 recently

-1.33

 they

-1.23

 últimas

-1.21

 każdy

-1.20

 vừa

-1.18

 ludzie

-1.16

 którzy

-1.13

-1.11

POSITIVE LOGITS

紲

1.16

by

1.16

 then

1.14

ariamente

1.13

reté

1.13

 lais

1.12

见证

1.09

 kantor

1.09

⋀

1.08

입니다

1.06

Activations Density 0.101%