INDEX

Explanations

fewer

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

 nothing

-1.61

nothing

-1.54

Nothing

-1.48

 Nothing

-1.40

 treating

-1.22

 treat

-1.20

 fewer

-1.16

NOTHING

-1.16

 NOTHING

-1.13

 ingenting

-1.09

POSITIVE LOGITS

EndGlobalSection

0.80

 متعلقه

0.77

 jsPsych

0.75

 himo

0.71

ValueStyle

0.69

enumi

0.67

enumii

0.66

/******/

0.66

Derbyniad

0.65

:✨

0.64

Activations Density 0.195%