INDEX

Explanations

listing technical, movie, and song titles

The main thing this neuron does is detect numeric tokens (i.e. standalone numbers).

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

If

-1.98

-1.91

 With

-1.88

 Each

-1.86

 There

-1.80

 Although

-1.69

 Once

-1.68

Use

-1.67

-1.65

 After

-1.63

POSITIVE LOGITS

があるので

2.23

がありますが

2.13

<bos>

2.06

毟

1.77

や

1.76

🫃

1.75

܆

1.73

などを

1.68

 nonché

1.67

휀

1.64

Activations Density 0.358%