INDEX

Explanations

source of information

This neuron detects the noun or noun phrase that follows the trigger phrase “source of.”

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

-1.85

用的

-1.42

 there

-1.34

 одной

-1.20

 одного

-1.18

 verschiedenen

-1.16

 mehreren

-1.13

 ponemos

-1.12

 onderdeel

-1.09

Ken

-1.08

POSITIVE LOGITS

for

1.70

and

1.53

 should

1.52

 both

1.46

 which

1.37

 that

1.31

 would

1.29

 namely

1.29

 intervento

1.27

gesamt

1.19

Activations Density 0.039%