INDEX

Explanations

references to assistant or assistant-related roles

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

〞

-0.66

 Schuld

-0.65

Kamu

-0.65

$_['

-0.64

genheim

-0.63

URIS

-0.63

 Zeug

-0.62

⁠

-0.60

 SafeMath

-0.59

etera

-0.59

POSITIVE LOGITS

 Asst

1.65

 Assn

1.51

 Univ

1.48

ppl

1.27

yrs

1.24

 govt

1.20

 Govt

1.20

Univ

1.13

 Dept

1.10

Mfg

1.10

Activations Density 0.021%