INDEX

Explanations

based on or generated from

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

 seldom

1.32

 rarely

1.31

 rarement

1.19

 quelquefois

1.16

 muchas

1.14

 größeren

1.13

 numerosas

1.11

 muitas

1.10

 hardly

1.09

 poderos

1.07

POSITIVE LOGITS

 centered

1.18

 centred

1.10

 reliant

1.09

 сосредото

1.08

 driven

0.99

reliant

0.97

 concentrated

0.93

 focused

0.93

focused

0.91

 stemming

0.91

Activations Density 0.237%