INDEX

Explanations

private [noun]

The neuron activates specifically on the adjective “private.”

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

氿

-1.92

 Não

-1.88

-1.82

}$

-1.80

</strong>

-1.73

蛲

-1.72

 vinte

-1.71

 into

-1.67

THAT

-1.67

-1.66

POSITIVE LOGITS

 навіть

2.31

Even

2.30

Surprisingly

2.23

Даже

2.22

Олег

2.19

᭝

2.17

Despite

2.14

Although

2.06

 間取り

2.03

ámicas

1.98

Activations Density 0.011%