INDEX

Explanations

skills shifts and gaps

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

ontiti

3.35

ição

3.24

楶

3.14

აწ

3.09

ালক

3.09

ações

3.07

तः

3.03

レル

3.02

लिसिस

3.02

uirre

2.98

POSITIVE LOGITS

ding

2.44

bin

2.42

ben

2.30

뭔

2.28

met

2.23

িয়ে

2.13

ermann

2.13

콧

2.13

wirk

2.13

맹

2.12

Activations Density 0.098%