INDEX

Explanations

starting or ending segments

This neuron detects any token containing the letter sequence “ope.”

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

｣

-2.28

֢

-2.09

and

-1.93

 новым

-1.91

 Würde

-1.88

 while

-1.83

 even

-1.82

楽しめます

-1.80

 have

-1.73

 time

-1.73

POSITIVE LOGITS

鹇

2.23

and

2.22

ments

2.09

了许多

2.06

ific

2.05

</strong>

2.05

usive

2.05

圌

2.05

ation

2.03

 historien

2.02

Activations Density 0.024%