INDEX

Explanations

types, styles, or genres

The neuron strongly activates on words that label or categorize things—i.e. terms like “type,” “category,” “style,” “format,” or “level.”

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

and

1.18

وت

1.13

AVE

1.13

ﻪ

1.07

ина

1.04

чал

1.03

ंग

1.00

भर

1.00

ast

0.97

OUTH

0.97

POSITIVE LOGITS

ভেদ

1.21

ilk

1.09

 oblik

1.09

 endeav

1.05

 wijze

1.03

 estilos

1.02

 echter

1.02

 এছাড়াও

1.02

 类型

1.00

 iler

1.00

Activations Density 3.450%