INDEX

Explanations

wasn't, isn't, wouldn't

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

ت

3.32

ために

3.05

ా

2.96

ি

2.91

eenth

2.84

𝙮

2.77

ی

2.76

ための

2.73

ੀ

2.67

HCR

2.64

POSITIVE LOGITS

не

2.53

ργαν

2.29

ImmediateContext

2.28

StoreException

2.18

狽

2.11

ра

2.09

不起

2.07

р

2.06

Tarea

2.04

nal

2.03

Activations Density 0.059%