INDEX

Explanations

words starting with "Ind"

The neuron is keyed to the character sequence “ind,” activating strongly on any token containing that trigram.

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

llama

-0.77

 kür

-0.69

Oni

-0.68

slowly

-0.68

bign

-0.67

 команде

-0.67

cmb

-0.66

巻

-0.66

ADA

-0.66

Emotion

-0.66

POSITIVE LOGITS

Ind

1.38

IND

1.35

Ind

1.30

indy

1.18

indi

1.16

indu

1.13

IND

1.13

ind

1.13

 Indi

1.09

idual

1.07

Activations Density 0.011%