INDEX

Explanations

recommendations and advice

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

 nét

-1.13

キャンセル

-0.98

"@/

-0.95

haku

-0.93

Приятного

-0.93

 JOSÉ

-0.93

欧盟

-0.92

 Girolamo

-0.92

⠟

-0.91

 Rabu

-0.91

POSITIVE LOGITS

for

1.91

 recommended

1.82

 recommend

1.79

 advice

1.74

 advisable

1.70

 recommendation

1.66

 advised

1.66

 should

1.65

you

1.63

 recommendations

1.61

Activations Density 0.052%