INDEX

Explanations

names followed by comma

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

</td>

-1.70

 возника

-1.59

<td>

-1.57

比如说

-1.55

ー

-1.52

</i>

-1.49

咵

-1.46

ut

-1.45

"));

-1.41

How

-1.39

POSITIVE LOGITS

but

1.65

these

1.59

 these

1.52

we

1.51

することも

1.45

閬

1.43

 this

1.42

 again

1.41

時は

1.39

 behulp

1.34

Activations Density 0.083%