INDEX

Explanations

mathematical and optimization concepts

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

tikai

-0.85

 ikiwa

-0.83

brado

-0.83

诒

-0.82

 njia

-0.80

anes

-0.78

 next

-0.78

ポーチ

-0.77

 Suggested

-0.77

包括

-0.75

POSITIVE LOGITS

 Removing

0.93

стера

0.86

GAT

0.86

 handelt

0.84

hadapi

0.83

Remove

0.83

allinen

0.83

 tekrar

0.82

МО

0.81

spk

0.81

Activations Density 0.063%