INDEX

Explanations

code snippets with method calls

This neuron strongly activates on numeric literal tokens (e.g., “69”, “18.625”, “54.75”)—i.e. it detects numbers in code.

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

to

-1.73

on

-1.63

in

-1.45

and

-1.42

uação

-1.41

 have

-1.41

=@"

-1.39

。“

-1.38

ToRefresh

-1.34

datable

-1.34

POSITIVE LOGITS

ってない

1.73

不夠

1.54

していない

1.53

Кстати

1.52

 visant

1.51

屢

1.49

也沒有

1.48

也不會

1.43

還會

1.41

並沒有

1.41

Activations Density 0.016%