INDEX

Explanations

extracting features

New Auto-Interp

Configuration

Prompts (Dashboard)

238,145 prompts, 512 tokens each

Dataset (Dashboard)

lmsys + oasst1

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 satisfiable

0.77

隐含规则

0.72

 decryption

0.69

 causality

0.69

 inference

0.67

しゃ

0.67

かう

0.66

 Continuity

0.66

 Goss

0.66

 അനു

0.66

POSITIVE LOGITS

 features

2.03

 Features

2.01

 feature

2.00

Feature

1.98

Features

1.92

 Feature

1.87

features

1.85

feature

1.79

 FEATURES

1.69

 FEATURE

1.48

Activations Density 0.119%