INDEX

Explanations

goals and project aims

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

＾

-1.02

ículos

-1.00

比べて

-0.90

仕組み

-0.89

każ

-0.84

 what

-0.83

Karol

-0.82

 astronomical

-0.81

 protokol

-0.78

いわゆる

-0.78

POSITIVE LOGITS

 Goal

1.27

 Objective

1.20

 requirement

1.13

 indicator

1.13

 goal

1.09

 item

1.09

 subgoal

1.09

 Criterion

1.05

 relacionada

1.03

 deliver

1.02

Activations Density 0.031%