INDEX

Explanations

clarity and usability

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

 kriy

1.04

 privilegio

1.03

 elastic

0.99

释放

0.98

Plas

0.96

 Elsa

0.96

 يوليو

0.95

 congel

0.94

 Pleistocene

0.94

 Bash

0.94

POSITIVE LOGITS

 guides

1.54

 guide

1.52

Guide

1.43

 guidance

1.42

 Guides

1.40

 Guide

1.40

Guides

1.40

識別

1.32

 navigation

1.29

guides

1.29

Activations Density 0.470%