INDEX

Explanations

quantified parameters and categories

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ｇ

1.13

 Carpath

1.13

 Berk

1.11

Lat

1.11

 Zeitschr

1.10

 yıldır

1.10

 Roth

1.09

 hundert

1.09

Lod

1.09

 Marg

1.08

POSITIVE LOGITS

 NOTE

1.19

Note

1.16

 تص

1.15

 PROCEDURE

1.14

fehler

1.13

 ابن

1.12

Assessment

1.09

 оце

1.08

Consider

1.08

чи

1.07

Activations Density 0.354%