INDEX

Explanations

saying or considering something

The neuron detects numeric tokens—any standalone numbers (including integers and decimals).

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

 или

-1.43

 både

-1.34

or

-1.25

 escritas

-1.23

UINTN

-1.22

to

-1.22

↵↵↵↵↵↵

-1.17

 یا

-1.15

 acquire

-1.13

 entweder

-1.13

POSITIVE LOGITS

</i>

1.56

drawSprites

1.25

 existência

1.23

 gestartet

1.18

 schwarzen

1.16

 německ

1.16

ские

1.16

Dette

1.15

niño

1.13

一個

1.12

Activations Density 0.042%