INDEX

Explanations

actor, ROW, Drupal, ECT

This neuron fires on tokens that are in (or begin with) uppercase‐letter sequences—e.g. all-caps keywords, acronyms, or identifiers and other tokens starting with a capital letter.

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

not

-1.94

new

-1.88

 with

-1.83

all

-1.80

 from

-1.77

 even

-1.68

 puts

-1.67

If

-1.67

 using

-1.64

 without

-1.64

POSITIVE LOGITS

 ส์

1.66

܆

1.59

 faciliter

1.56

 rapporter

1.50

 relâche

1.42

ఽ

1.42

 Dinas

1.41

 essas

1.40

 acheter

1.38

 AMAZING

1.37

Activations Density 0.006%