INDEX

Explanations

two followed by descriptors

The neuron detects occurrences of the word “two” used to enumerate or contrast a pair of items.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

it

-1.52

。(

-1.37

！（

-1.34

或者

-1.30

 Siehe

-1.30

 acque

-1.26

 سایر

-1.23

at

-1.22

 vedere

-1.16

！！！

-1.15

POSITIVE LOGITS

 этих

1.85

 contemporáneo

1.59

 extremely

1.59

都不能

1.58

糨

1.56

 wächst

1.55

 estudian

1.50

 exceedingly

1.49

違って

1.46



1.41

Activations Density 0.023%