INDEX

Explanations

numbers followed by quantities

The neuron strongly activates on numeric tokens—particularly numbers, decimals, and percentages.

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



-1.61

 What

-1.47

 немає

-1.41

as

-1.38

 Commencez

-1.36

 with

-1.36

 publiés

-1.35

 značky

-1.34

Hispanic

-1.34

४

-1.34

POSITIVE LOGITS

of

2.36

された

1.65

则

1.54

ardige

1.52

 ſtand

1.46

ですね

1.42

 enceinte

1.41

 occaf

1.41

 cilvē

1.40

 heridos

1.39

Activations Density 0.070%