INDEX

Explanations

awards and titles

The neuron activates on ordinal numbering within formal titles—e.g. “4th International Conference,” “63rd Indian Infantry Brigade,” “47th British Academy Film Awards,” “37th studio album,” 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

 概要

-1.18

 recente

-1.17

就可以了

-1.13

 récente

-1.12

 nieuwe

-1.06

了过来

-1.01

 Pilate

-1.01

很赞

-1.00

就行了

-1.00

 isim

-0.98

POSITIVE LOGITS

in

1.25

 since

1.16

1.06

of

1.05

と言われる

1.01

 tersebut

0.96

 Since

0.93

 While

0.92

 tradition

0.92

 while

0.90

Activations Density 0.072%