INDEX

Explanations

descriptions and explanations

The neuron fires on standalone headings or titles (e.g. section names, product names, page titles) in the text.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

KeyCode

0.71

USH

0.70

TypeDef

0.70

GHz

0.67

NE

0.66

𝚝

0.63

 EtOAc

0.63

AE

0.62

 परेशानी

0.62

0.61

POSITIVE LOGITS

 informativo

1.38

 informativa

1.30

 display

1.23

 displaying

1.22

 information

1.20

 summaries

1.15

 informasi

1.14

 informazioni

1.14

 información

1.13

 descriptions

1.13

Activations Density 2.809%