INDEX

Explanations

Hill, entitled, shepherd

setting following `llama`

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

راط

0.46

用

0.45

งาน

0.43

ㅅ

0.43

бля

0.43

tać

0.42

 laboratoires

0.41

оте

0.41

ەڵ

0.40

каз

0.40

POSITIVE LOGITS

 hilly

0.56

 alleviating

0.51

ruciating

0.49

ETING

0.49

 kurta

0.48

ining

0.48

 প্রচ

0.47

 তৎকালীন

0.47

 keratin

0.46

adihi

0.46

Activations Density 0.000%