INDEX

Explanations

required or responsible

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

asm

-0.11

inan

-0.11

emo

-0.11

 parity

-0.10

 Wand

-0.10

ybrid

-0.10

folk

-0.10

_reason

-0.10

 duly

-0.10

vore

-0.10

POSITIVE LOGITS

 need

0.14

 must

0.14

 MUST

0.11

 automatically

0.11

å¿ħ

0.10

must

0.10

 Must

0.10

 always

0.10

 gotta

0.10

 required

0.10

Activations Density 0.367%