INDEX

Explanations

in

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

DebuggerNonUser

-0.77

JsonKey

-0.68

GraphicsUnit

-0.65

principalTable

-0.64

(!__

-0.63

mtrl

-0.63

 незавершена

-0.63

 nakalista

-0.63

 writing

-0.61

Portale

-0.61

POSITIVE LOGITS

ksen

0.47

 complètes

0.47

ksessa

0.47

 Polsek

0.47

/$',

0.45

 חיצוניים

0.45

 misura

0.45

 religieuses

0.43

<bos>

0.43

原标题

0.41

Activations Density 0.026%