INDEX

Explanations

research results

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 results

-0.60

<eos>

-0.54

↵↵

-0.49

 výsled

-0.48

-0.47

Results

-0.47

↵

-0.47

results

-0.46

の結果

-0.45

POSITIVE LOGITS

Efq

1.14

تقاوى

1.13

 Majefty

1.11

WithIOException

1.11

Бахар

1.09

complexContent

1.05

 auroit

1.03

IContainer

1.01

 ainfi

1.01

URLException

1.00

Activations Density 0.031%