INDEX

Explanations

fetch retrieve

information retrieval

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

屈

0.45

畊

0.42

ель

0.41

Ending

0.41

дій

0.40

 gören

0.40

转向

0.38

BackgroundHelper

0.38

Carroll

0.38

 pozost

0.38

POSITIVE LOGITS

 Fetch

1.39

 Retrieve

1.36

 fetch

1.35

 Content

1.33

 content

1.30

 retrieve

1.29

 FETCH

1.24

Fetch

1.21

retrieve

1.21

FETCH

1.20

Activations Density 0.005%