INDEX

Explanations

is a noun/adjective/participle

The neuron strongly activates on tokens that are three-digit numbers (i.e. large numeric quantities).

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

 then

-1.77

or

-1.53

 विभिन्न

-1.41

 such

-1.38

edere

-1.35

badan

-1.33

 nimmer

-1.28

👍👍

-1.23

 these

-1.21

teeth

-1.21

POSITIVE LOGITS

 already

1.29

ovico

1.27

was

1.27

わずか

1.26

久しぶり

1.26

使った

1.25

一趟

1.23

龊

1.23

 damos

1.22

indakan

1.22

Activations Density 0.076%