INDEX

Explanations

outcomes of experiments

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

-2.13

-2.11

 включает

-2.09

 где

-2.06

有一

-2.03

 Для

-1.95

 получить

-1.95

 moest

-1.94

 discomfort

-1.92

POSITIVE LOGITS

 results

3.13

葙

2.48

歺

2.45

㈰

2.44

噚

2.39

 纹身

2.28

ﮯ

2.27

 Méri

2.23

 Presiden

2.22

 中式

2.20

Activations Density 0.028%