INDEX

Explanations

references to Torah

The neuron activates on the word “Torah” (i.e. it detects mentions of the Torah).

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

 цього

-2.56



-2.50

-2.48

 unſ

-2.44

というわけで

-2.38

-2.34

 greateſt

-2.31

什么的

-2.31

CurtirCurtir

-2.27

-2.23

POSITIVE LOGITS

2.84

蜴

2.38

讹

2.30

to

2.25

mathrm

2.19

 tidligere

2.17

鉭

2.17

 they

2.17

ebbe

2.14

瘣

2.13

Activations Density 0.001%