INDEX

Explanations

built/made for/from

The neuron fires most strongly on words that describe how something is produced or designed—i.e. manufacturing/production verbs (made, designed, cast, produces, etc.).

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

 issuers

0.54

्वी

0.53

 occurring

0.52

 हनु

0.52

 égard

0.52

 adduced

0.52

 निर्णया

0.51

 diperoleh

0.50

 issued

0.50

 düzenlenen

0.50

POSITIVE LOGITS

 using

1.42

 باستخدام

1.37

using

1.29

 usando

1.24

 utilizzando

1.23

 используя

1.20

 utilizando

1.18

 USING

1.09

 Using

1.05

 menggunakan

1.04

Activations Density 0.654%