INDEX

Explanations

the word "of" in math problems

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Efq

-1.04

 سكانية

-1.02

তথ্যসূত্র

-0.95

)•

-0.94

)*/

-0.93

 itſelf

-0.91

 Forumite

-0.85

 myſelf

-0.85

 Majefty

-0.85

ſelf

-0.84

POSITIVE LOGITS

<eos>

0.51

↵↵

0.45

 when

0.43

PagerAdapter

0.42

dy

0.42

 near

0.42

0.41

hesis

0.41

↵

0.40

Activations Density 0.000%