INDEX

Explanations

explaining purpose or action

The neuron activates on standalone multi‐digit numerals—especially four‐digit numbers like years or large counts—in the text.

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

れません

-1.36

 paramètres

-1.20

to

-1.18

ʎ

-1.16

:''

-1.15

tront

-1.15

है

-1.15

 oignon

-1.14

טרה

-1.13

などは

-1.13

POSITIVE LOGITS

 help

2.41

 ensure

2.22

be

2.02

 avoid

1.88

 allow

1.70

get

1.66

はこんな感じ

1.46

try

1.44

 make

1.43

 which

1.41

Activations Density 0.210%