INDEX

Explanations

decision task

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

 Lala

1.20

Templ

1.20

 Templ

1.17

 naziv

1.17

 newcom

1.16

TÜ

1.13

şt

1.13

peza

1.12

罩

1.12

ihar

1.11

POSITIVE LOGITS

 decisions

3.08

 Decisions

3.00

 decision

2.86

decision

2.83

 Decision

2.68

 decisiones

2.62

Decision

2.59

 choices

2.57

 decisione

2.52

 decisão

2.52

Activations Density 1.069%