INDEX

Explanations

Ivan followed by other names

The neuron mainly picks out occurrences of the name “Ivan” (including as a substring in larger tokens such as “Ivanka”).

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

↵

-5.13

’

-4.63

-3.73

↵↵

-3.16

-2.86

of

-2.77

-2.70

,"

-2.55

",

-2.52

Of

-2.50

POSITIVE LOGITS

 universelle

2.88

錞

2.86

喾

2.84

嗐

2.81

ｏ

2.75

鲼

2.67

 Terme

2.66

HasIndex

2.66

櫨

2.64

lawesi

2.58

Activations Density 0.004%