INDEX

Explanations

prefixes to specific concepts

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

：

1.38

1.28

:$

1.24

1.19

めっちゃ

1.18

:"

1.18

->

1.17

():

1.12

後面

1.12

__:

1.10

POSITIVE LOGITS

 Furthermore

2.05

Furthermore

2.02

 Additionally

2.01

Additionally

1.99

 conversely

1.93

Contrary

1.92

 fluctuations

1.92

 Moreover

1.89

Alternatively

1.86

Moreover

1.86

Activations Density 0.271%