INDEX

Explanations

Xin names and entities

The neuron activates on any token containing the letter sequence “xin,” regardless of case or context.

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

</sup>

-2.66

”,

-2.56

It

-2.53

 ähnliche

-2.48

',

-2.42

 endereco

-2.34

 fazendo

-2.30

↵

-2.28

-2.25

You

-2.22

POSITIVE LOGITS

⦊

2.83

鏵

2.64

ate

2.45

can

2.41

ﮕ

2.33

there

2.25

👹

2.17

郤

2.16

 اجتماعی

2.11

ፃ

2.05

Activations Density 0.005%