INDEX

Explanations

fine-tuning with Greek or Slavic words

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

 croire

1.17

 egun

1.11

 craindre

1.07

 izan

1.06

longrightarrow

1.06

 revenu

1.06

 同じ

1.04

の結果

1.02

educt

0.97

पुढे

0.97

POSITIVE LOGITS

ích

1.53

ří

1.42

ští

1.41

ány

1.32

íc

1.31

má

1.31

rá

1.31

ých

1.31

rů

1.30

zené

1.29

Activations Density 0.247%