INDEX

Explanations

reacting or responding

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

頗

1.56

 Considerable

1.33

Incidentally

1.33

 agak

1.31

得很

1.30

比较

1.29

很不

1.29

 Incidentally

1.28

 Generally

1.28

 considerably

1.27

POSITIVE LOGITS

 reimag

1.77

 resonates

1.66

 curated

1.51

 aligns

1.47

🤍

1.47

 relatable

1.36

 curation

1.36

 ❤️

1.32

 resonate

1.32

 repurposed

1.31

Activations Density 0.202%