INDEX

Explanations

column names/features

This neuron responds to tokens that are part of tabular data—especially numeric literals and column‐header labels in DataFrame-style outputs.

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

脔

-1.51

our

-1.41

 いい

-1.35

你和

-1.35

摀

-1.34

谗

-1.32

今月

-1.31

 primeiros

-1.27

<<<<<<<<<<<<

-1.27

レーター

-1.26

POSITIVE LOGITS

edoria

1.54

訫

1.52

1.44

Conclusão

1.43

 tchèque

1.22

Outras

1.20

1.18

SECOND

1.17

1.16

除非

1.14

Activations Density 0.004%