INDEX

Explanations

languages

This neuron chiefly detects first-person plural marketing or promotional statements—i.e. sentences beginning with “We” (or “We” plus a promotional verb) that introduce company claims or service pitches.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 পরবর্তীতে

1.55

非常的

1.46

 oftentimes

1.45

 தற்போது

1.37

 oldukça

1.33

 देखील

1.31

 काफी

1.28

utilizzo

1.27

 بشكل

1.23

 শুধুমাত্র

1.21

POSITIVE LOGITS

 wept

1.03

 öld

0.95

 треба

0.94

 nahin

0.91

 不能

0.88

 不是

0.87

 trois

0.84

 killed

0.83

 Надо

0.83

 veremos

0.82

Activations Density 1.588%