INDEX

Explanations

instead of, rather than

This neuron activates on numeric tokens (multi-digit numbers) 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

and

-1.61

 الأمريكي

-1.54

这一幕

-1.53

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

-1.51

ÉD

-1.46

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

-1.46

dalam

-1.45

"¿

-1.45

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

-1.45

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

-1.45

POSITIVE LOGITS

 being

2.14

 have

1.93

只是

1.82

 merely

1.80

It

1.79

 just

1.69

1.62

1.59

 simply

1.59

it

1.58

Activations Density 0.014%