INDEX

Explanations

code delimiters or structure

The neuron activates on tokens that are purely numeric (i.e. sequences of digits).

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

 gainera

-1.15

 étanche

-1.10

 everyone

-1.09

 valla

-1.07

頷

-1.05

 rios

-1.04

 fhort

-1.03

ENZA

-1.02

'}';

-1.01

SIZES

-1.01

POSITIVE LOGITS

if

1.40

延伸閱讀

1.26

 грамм

1.23

を探す

1.13

 jeśli

1.13

Если

1.09

vaccinated

1.09

 emporter

1.05

 вокруг

1.03

 jeżeli

1.01

Activations Density 0.001%