INDEX

Explanations

refusal of harmful information

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

 Bent

0.90

角度

0.89

POV

0.88

돌

0.87

POV

0.86

 ângulo

0.85

 momentary

0.85

 góc

0.84

 زاويه

0.84

 angles

0.83

POSITIVE LOGITS

xaxis

1.21

yaxis

0.99

Paper

0.98

 Paper

0.98

紙

0.95

 xaxis

0.93

uum

0.93

xas

0.91

 शोक

0.90

 PAPER

0.89

Activations Density 0.060%