INDEX

Explanations

lastly concluding

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

回戦

-0.96

marzo

-0.94

स्

-0.94

Rationale

-0.90

ayangkan

-0.88

not

-0.86

 Placer

-0.86

 कभी

-0.84

Рис

-0.83

 reen

-0.83

POSITIVE LOGITS

 concluding

1.30

Lastly

1.30

 conclude

1.20

最後に

1.20

 hopefully

1.19

 concluyó

1.19

we

1.17

 lastly

1.12

 můžete

1.11

最后

1.09

Activations Density 0.002%