INDEX

Explanations

game results

This neuron detects numbers and numeric score patterns (e.g. “1-0,” “4–0,” decimal stats, etc.) 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

躂

-1.55

栳

-1.53

*******

-1.51

 групп

-1.48

)=>{

-1.47

 Allí

-1.46

ญี่ป

-1.44

ถ้า

-1.41

 Tentang

-1.40

 Quý

-1.39

POSITIVE LOGITS

by

2.48

for

1.90

ing

1.80

1.67

to

1.61

er

1.59

 also

1.50

 because

1.46

ar

1.44

 that

1.40

Activations Density 0.014%