INDEX

Explanations

military service or personnel

The neuron activates on words naming military branches, units, or service (e.g. Army, Marine Corps, Navy, Air Force, boot camp), i.e. references to U.S. military service.

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

久しぶり

-1.13

ごとの

-1.07

うまい

-1.07

なかな

-1.05

 queste

-1.05

の方法

-1.05

ところに

-1.05

イケメン

-1.04

の内容

-1.02

ならでは

-1.02

POSITIVE LOGITS

 where

1.46

 what

1.35

 during

1.29

 after

1.08

したいと思います

1.05

haviours

1.02

에서

1.00

 even

0.98

issä

0.96

where

0.96

Activations Density 0.019%