INDEX

Explanations

printing text and output

This neuron activates on function names used for console I/O (e.g., print, println, input, expect).

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

Präsident

-1.30

 PUSTAKA

-1.25

any

-1.22

Jahre

-1.20

 their

-1.19

vejec

-1.16

 grossesse

-1.16

źródło

-1.16

któ

-1.15

Wszyst

-1.14

POSITIVE LOGITS

1.86

 mentre

1.57

1.46

1.35

ha

1.34

is

1.34

''

1.30

an

1.28

on

1.26

1.24

Activations Density 0.003%