INDEX

Explanations

optional or placeholders like [insert]

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

↵↵

2.30

<h3>

2.06

<h2>

2.01

<eos>

1.78

<start_of_image>

1.76

<h4>

1.49

</h3>

1.42

"""

1.38

1.34

</td>

1.32

POSITIVE LOGITS

 admittedly

2.78

especially

2.75

특히

2.71

 кстати

2.68

particularly

2.67

 notably

2.63

 caveat

2.59

 doubtless

2.58

mittedly

2.58

 특히

2.52

Activations Density 0.283%