INDEX

Explanations

database queries and code

The neuron fires on PHP variable tokens, i.e. the “$” symbol indicating a variable.

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

 elevación

-1.61

:@"

-1.47

！”

-1.44

Française

-1.42

());

-1.42

 efectuar

-1.41

Comté

-1.41

čil

-1.39

brtc

-1.38

 nooit

-1.37

POSITIVE LOGITS

1.90

 TAMBÉM

1.63

1.62

1.46

等我

1.45

憮

1.44

 beira

1.42

 silêncio

1.40

アウトドア

1.40

よりも

1.39

Activations Density 0.004%