INDEX

Explanations

The neuron fires on dotted member‐access expressions (object.property or method calls using “.”).

scope

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

of

-1.05

and

-0.96

specific

-0.95

зира

-0.91

 just

-0.91

 simply

-0.91



-0.90

in

-0.89

on

-0.88

 mere

-0.88

POSITIVE LOGITS

╧

1.07

 مانت

1.00

 condado

0.95

晶

0.94

saturday

0.93

oscill

0.92

undred

0.90

 hvorfor

0.90

＝＝

0.89

 Prueba

0.89

Activations Density 0.008%