INDEX

Explanations

past participles

The neuron detects past-participle verbs used in “has/have been” passive constructions.

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

 последний

-1.16

的時候

-1.14

 ‘‘

-1.13

 जब

-1.13

最后

-1.09

 одной

-1.07

 still

-1.04

有没有

-1.04

 schafft

-1.03

 ricord

-1.02

POSITIVE LOGITS

 since

2.63

 lately

2.00

since

1.78

to

1.65

 recientemente

1.60

 been

1.59

for

1.50

 recentemente

1.49

 SINCE

1.48

 Lately

1.41

Activations Density 0.079%