INDEX

Explanations

the word "candidate", sometimes followed by "for."

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

綜合報導

-0.57

wirkung

-0.56

MemoryWarning

-0.56

 Habit

-0.53

 supervision

-0.52

STO

-0.51

 protectors

-0.51

`,

-0.51

）』

-0.51

 Gine

-0.51

POSITIVE LOGITS

 candidates

2.92

 candidate

2.89

 Candidates

2.67

 Candidate

2.63

candidate

2.58

candidates

2.53

Candidate

2.52

Candidates

2.52

 CANDIDATE

2.42

 candidatos

2.20

Activations Density 0.443%