INDEX

Explanations

app descriptions

The neuron fires on the standalone token “app,” i.e. mentions of applications.

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

It

-2.55

而

-2.39

 オーバー

-2.38

僰

-2.34

☫

-2.28

驁

-2.28

 potes

-2.20

IBLIO

-2.20

があって

-2.19

 movimenta

-2.16

POSITIVE LOGITS

2.88

ୈ

2.64

 ......

2.59

 buchen

2.58

 prenotazione

2.52

2.50

2.48

著名的

2.42

 .....

2.41

Getting

2.39

Activations Density 0.048%