INDEX

Explanations

affected by

The neuron detects past-participial verbs used in passive constructions—verbs like “touched,” “affected,” or “devastated” occurring with “by” to mark being impacted by something.

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

谂

-1.48

مه

-1.35

 indisponible

-1.20

ativo

-1.17

 угро

-1.17

 ſta

-1.16

owaniem

-1.14

 темпера

-1.13

зраст

-1.13

து

-1.13

POSITIVE LOGITS

 these

1.66

 those

1.60

 تلك

1.52

 with

1.49

and

1.48

 that

1.27

又は

1.25

들에게

1.21

 bahwa

1.21

樫

1.20

Activations Density 0.045%