INDEX

Explanations

foreign language fragments

The neuron specifically detects country names (toponyms) in various languages and scripts.

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

-2.33

！"

-2.19

 from

-2.06

 because

-2.02

！”

-1.98

："

-1.98

：@

-1.93

 навіть

-1.89

！！”

-1.86

มากๆ

-1.81

POSITIVE LOGITS

</h4>

2.53

ਃ

2.36

 suena

1.97

</b>

1.91

냑

1.88

 остальные

1.85

 några

1.81

颯

1.77

 Ensuring

1.73

威胁

1.72

Activations Density 0.048%