INDEX

Explanations

technical writing terms, most notably "tractable"

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

riwal

-0.75

+#+

-0.73

enumii

-0.72

lorette

-0.69

AccessorTable

-0.65

 Insee

-0.65

 propOrder

-0.65

mbic

-0.65

 crdi

-0.65

#+#

-0.64

POSITIVE LOGITS

Mar

0.50

 easy

0.50

sub

0.47

easy

0.42

rõ

0.42

LabelTagHelper

0.42

Mar

0.42

 للمعارف

0.42

 rest

0.41

 beho

0.40

Activations Density 0.178%