INDEX

Explanations

Gupta surnames

The neuron is strongly activated by personal surnames (especially author last names in citations).

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

銠

-3.06

玉ねぎ

-2.84

 They

-2.83

馘

-2.80

甡

-2.78

ༀ

-2.75

ᕱ

-2.69

鴆

-2.58

);

-2.55

鸶

-2.47

POSITIVE LOGITS

 When

3.09

 they

2.86

釭

2.81

</h3>

2.69

 diğer

2.59

 Additionally

2.56

 Conversely

2.53

 Мы

2.52

々は

2.44

 เพื่อ

2.42

Activations Density 0.000%