INDEX

Explanations

describing individuals' unique traits

The neuron activates on formal evaluative phrases of the form “qualify as a ‘fine institution,’” i.e. when someone or something is described as qualifying as an institution.

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

it

-1.31

its

-1.07

 оно

-1.04

 которое

-0.98

 มัน

-0.95

 which

-0.95

它的

-0.95

它

-0.94

amek

-0.88

риях

-0.87

POSITIVE LOGITS

 himself

2.13

himself

1.62

 person

1.59

who

1.55

 whom

1.53

 incarnate

1.24

 figure

1.24

 Himself

1.23

 personnage

1.22

 embodiment

1.19

Activations Density 0.051%