INDEX

Explanations

imagine opening scenarios

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 hoped

-0.11

+'

-0.11

[Index

-0.11

 Wend

-0.11

ury

-0.10

 onward

-0.10

agem

-0.10

IPA

-0.10

 Schwarz

-0.10

stdin

-0.10

POSITIVE LOGITS

Imagine

0.17

 Imagine

0.16

 yourself

0.15

 imagine

0.12

 effortlessly

0.12

 seamlessly

0.12

sip

0.10

 your

0.10

 waking

0.10

 suddenly

0.10

Activations Density 0.034%