INDEX

Explanations

python pandas

The neuron detects the “import pandas as pd” pattern in code.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

⚋

-1.39

 一人

-1.34

 will

-1.34

at

-1.28

copic

-1.28

-1.27

THE

-1.26

 知道

-1.25

 bunun

-1.23

POSITIVE LOGITS

 egenskaper

1.76

 ansik

1.73

 sosis

1.60

merking

1.57

 boken

1.54

 teater

1.52

 eſſ

1.48

 rajut

1.48

 ANGLE

1.48

我們

1.47

Activations Density 0.016%