INDEX

Explanations

you want to

The neuron responds to second‐person address in the text—i.e. occurrences of “you” (and related “your”) directly speaking to the reader.

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

 работо

-0.74

 negocio

-0.72

ريكا

-0.70

Rapid

-0.69

aniem

-0.69

 دن

-0.69

 DEBUG

-0.69

anv

-0.68

thage

-0.68

Ont

-0.67

POSITIVE LOGITS

 simply

1.27

 simplesmente

0.86

simply

0.85

 lacquered

0.80

 simplement

0.79

Simply

0.77

 jät

0.75

 Simply

0.75

 grø

0.73

atering

0.69

Activations Density 0.075%