INDEX

Explanations

following steps

The neuron fires on numbered step or section markers (e.g. “Step 2,” “3.”, “Method,” etc.).

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

 these

-2.09

 those

-1.72

 also

-1.42

These

-1.36

Also

-1.31

вими

-1.20

 both

-1.17

our

-1.17

 kupa

-1.17

 kveld

-1.16

POSITIVE LOGITS

 then

3.66

 Then

2.00

then

1.95

 затем

1.93

然后

1.93

Then

1.88

 kemudian

1.82

 następnie

1.81

 سپس

1.77

 thereafter

1.73

Activations Density 0.011%