INDEX

Explanations

The neuron fires on numerical information—digits or number words denoting quantities, counts, or stats.

is or are followed by past participle

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

-1.35

 which

-1.32

 with

-1.27

but

-1.23

という

-1.18

や

-1.17

to

-1.10

ines

-1.09

そして

-1.08

new

-1.07

POSITIVE LOGITS

 leveren

1.41

のですか

1.38

んですね

1.31

感じが

1.30

のですね

1.30

感じで

1.27

際には

1.25

 bekijken

1.25

ようになって

1.24

ものに

1.22

Activations Density 0.164%