INDEX

Explanations

Paris or Parisian

The neuron specifically detects mentions of the city name “Paris.”

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

--

-2.25

 unſ

-2.00

<bos>

-1.98

↵↵

-1.95

 могут

-1.93

而非

-1.88

.”

-1.85

")

-1.85

 суммы

-1.84

 wuß

-1.83

POSITIVE LOGITS



2.14

of

2.08

}$

1.98

 parfüm

1.85



1.84

壜

1.83

 alors

1.81

 Such

1.78

 betek

1.77

 zowel

1.75

Activations Density 0.008%