INDEX

Explanations

fill-in-the-blank questions

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

LLO

2.23

∈</

2.19

ANDEZ

2.12

LEMN

2.11

Sitting

2.08

வரையும்

2.08

VISA

2.02

WRAP

2.00

ווע

1.99

CTIONS

1.98

POSITIVE LOGITS

es

1.89

egi

1.87

 fase

1.85

urs

1.82

ilk

1.79

ळा

1.77

ائق

1.73

akati

1.71

ة

1.69

querda

1.67

Activations Density 0.025%