INDEX

Explanations

code and explanations

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

 Konz

-0.72

 رفت

-0.69

 follow

-0.68

ளை

-0.68

 receive

-0.68

 Prie

-0.67

Inverted

-0.67

させていただ

-0.67

 Kult

-0.66

imbo

-0.66

POSITIVE LOGITS

(*

0.99

let

0.94

let

0.93

Printf

0.84

*)

0.84

indt

0.83

 reopened

0.83

oca

0.83

 dune

0.81

 functor

0.79

Activations Density 0.040%