INDEX

Explanations

HTML opening tags, especially div tags, often with an activation value corresponding to the nesting depth

<

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

Negative Logits

so

-0.57

emptory

-0.55

-0.51

gnet

-0.50

Sin

-0.49

ev

-0.48

Bel

-0.47

 sold

-0.47

POSITIVE LOGITS

 pleaſure

1.20

 itſelf

1.19

 myſelf

1.16

 Majefty

1.04

Efq

1.01

 themſelves

0.99

ſelf

0.98

 Reſ

0.98

ſelves

0.96

 Houſe

0.96

Activations Density 0.136%