INDEX

Explanations

vehicle model years

This neuron never activates—it does not detect any pattern in the input.

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

がありますが

-1.36

 TÉCNICA

-1.34

 NATIONS

-1.33

もありました

-1.27

 TÉCN

-1.27

よー

-1.24

子は

-1.23

ｿ

-1.22

estões

-1.21

ebly

-1.20

POSITIVE LOGITS

為に

1.07

 jota

1.05

 следова

1.04

1.03

1.02

 државе

1.02

癱

1.00

How

0.98

Activations Density 0.021%