INDEX

Explanations

list item markers

The neuron activates strongly on HTML/XML tag syntax (angle brackets and tag names like `<a>`, `<ul>`, `<div>`, etc.).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 etik

-1.41

￢

-1.27

 चीज़ों

-1.19

ysteine

-1.18

şem

-1.16

 kutu

-1.12

 doulou

-1.12

ICKS

-1.12

 CGRectMake

-1.11

кация

-1.11

POSITIVE LOGITS

等は

1.46

These

1.35

According

1.27

according

1.23

 ditentukan

1.22

1.16

 genaue

1.13

apprendre

1.13

 такі

1.11

 angegebenen

1.10

Activations Density 0.076%