INDEX

Explanations

sentence start words

The neuron activates on standalone numeric tokens (i.e. sequences of digits).

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

Now

-1.68

:],

-1.67

al

-1.59

 exquisite

-1.54

<blockquote>

-1.53

 While

-1.52

𖥻

-1.52

 همچ

-1.51

 this

-1.49

 maintains

-1.47

POSITIVE LOGITS

1.74

他们

1.74

1.73

䨋

1.63

</b>

1.59

،

1.57

1.55

隳

1.50

1.49

 เก

1.48

Activations Density 0.160%