INDEX

Explanations

words and phrases associated with forming a hypothesis or analyzing data.

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

alamu

-0.90

 Multi

-0.68

Multi

-0.66

 multi

-0.65

 якості

-0.65

-0.62

popd

-0.60

MultipartFile

-0.60

 Beat

-0.60

lotte

-0.59

POSITIVE LOGITS

 claims

0.97

 rằng

0.97

 believed

0.93

 believes

0.93

 believe

0.91

 suggest

0.91

 claim

0.89

andExpect

0.89

 Claims

0.88

 Sugges

0.88

Activations Density 8.412%