INDEX

Explanations

some stickers

The neuron strongly activates on the indefinite quantifier “some.”

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

ामध्ये

1.48

ść

1.28

ння

1.23

ᆸ

1.23

ション

1.22

ामुळे

1.21

ောင့်

1.20

 nombreuses

1.19

ről

1.18

সমাজ

1.18

POSITIVE LOGITS

bodies

1.60

 الشيء

1.60

ones

1.46

מ

1.42

THING

1.25

 semblance

1.17

1.16

how

1.16

ară

1.16

Activations Density 0.173%