INDEX

Explanations

implications and outcomes

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

原标题

-0.72

Ast

-0.71

PathVariable

-0.70

 Porsche

-0.69

omina

-0.69

には

-0.68

between

-0.68

fag

-0.67

ンク

-0.66

Between

-0.66

POSITIVE LOGITS

 flows

3.25

 flowing

3.19

 flowed

3.03

 flow

3.00

 arising

2.98

flow

2.77

Flow

2.70

 wynika

2.66

 Flow

2.63

flowing

2.63

Activations Density 0.092%