INDEX

Explanations

origin or event context

The neuron strongly activates on numeric measurements and related units (e.g. ages, decimals, counts) in the text.

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

 pengar

-1.11

calling

-1.08

 inutile

-1.04

 suna

-1.02

 confortable

-1.02

 nepal

-1.01

 citroen

-1.01

 الآن

-1.00

 meeste

-0.98

 geforce

-0.98

POSITIVE LOGITS

 during

1.63

by

1.36

use

1.33

one

1.27

 when

1.17

 after

1.16

During

1.16

 through

1.15

 while

1.14

 accidentally

1.13

Activations Density 0.108%