Model: SoLU Model: 12 Layers, 6144 Neurons per Layer

Dataset: 80% C4 (Web Text) and 20% Python Code

Neuron 3465 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-12l')

Text #0

Max Range: 4.0036. Min Range: -4.0036

Max Act: 4.0036. Min Act: -0.0001

Data Index: 942221 (C4 (Web Text))

Max Activating Token Index: 60

Text #1

Max Range: 4.0036. Min Range: -4.0036

Max Act: 3.8748. Min Act: -0.0000

Data Index: 287201 (C4 (Web Text))

Max Activating Token Index: 460

Text #2

Max Range: 4.0036. Min Range: -4.0036

Max Act: 3.3056. Min Act: -0.0000

Data Index: 490210 (C4 (Web Text))

Max Activating Token Index: 801

Text #3

Max Range: 4.0036. Min Range: -4.0036

Max Act: 3.3092. Min Act: -0.0000

Data Index: 1297944 (C4 (Web Text))

Max Activating Token Index: 851

Text #4

Max Range: 4.0036. Min Range: -4.0036

Max Act: 3.2646. Min Act: -0.0001

Data Index: 652144 (C4 (Web Text))

Max Activating Token Index: 95

Text #5

Max Range: 4.0036. Min Range: -4.0036

Max Act: 2.9414. Min Act: -0.0001

Data Index: 809774 (C4 (Web Text))

Max Activating Token Index: 404

Text #6

Max Range: 4.0036. Min Range: -4.0036

Max Act: 3.4185. Min Act: -0.0000

Data Index: 1098315 (C4 (Web Text))

Max Activating Token Index: 50

Text #7

Max Range: 4.0036. Min Range: -4.0036

Max Act: 3.1990. Min Act: -0.0000

Data Index: 1628809 (Python Code)

Max Activating Token Index: 765

Text #8

Max Range: 4.0036. Min Range: -4.0036

Max Act: 3.0425. Min Act: -0.0001

Data Index: 1459513 (Python Code)

Max Activating Token Index: 644

Text #9

Max Range: 4.0036. Min Range: -4.0036

Max Act: 3.1811. Min Act: -0.0001

Data Index: 886857 (C4 (Web Text))

Max Activating Token Index: 123

Text #10

Max Range: 4.0036. Min Range: -4.0036

Max Act: 3.1966. Min Act: -0.0001

Data Index: 1607912 (Python Code)

Max Activating Token Index: 1010

Text #11

Max Range: 4.0036. Min Range: -4.0036

Max Act: 3.0170. Min Act: -0.0001

Data Index: 609781 (C4 (Web Text))

Max Activating Token Index: 639

Text #12

Max Range: 4.0036. Min Range: -4.0036

Max Act: 3.1450. Min Act: -0.0000

Data Index: 295725 (C4 (Web Text))

Max Activating Token Index: 1014

Text #13

Max Range: 4.0036. Min Range: -4.0036

Max Act: 2.9415. Min Act: -0.0001

Data Index: 428238 (C4 (Web Text))

Max Activating Token Index: 795

Text #14

Max Range: 4.0036. Min Range: -4.0036

Max Act: 2.9478. Min Act: -0.0001

Data Index: 75028 (C4 (Web Text))

Max Activating Token Index: 98

Text #15

Max Range: 4.0036. Min Range: -4.0036

Max Act: 3.1818. Min Act: -0.0001

Data Index: 1399205 (Python Code)

Max Activating Token Index: 806

Text #16

Max Range: 4.0036. Min Range: -4.0036

Max Act: 3.3118. Min Act: -0.0001

Data Index: 1550821 (Python Code)

Max Activating Token Index: 948

Text #17

Max Range: 4.0036. Min Range: -4.0036

Max Act: 3.3363. Min Act: -0.0001

Data Index: 1546414 (Python Code)

Max Activating Token Index: 720

Text #18

Max Range: 4.0036. Min Range: -4.0036

Max Act: 2.9173. Min Act: -0.0001

Data Index: 1176858 (C4 (Web Text))

Max Activating Token Index: 161

Text #19

Max Range: 4.0036. Min Range: -4.0036

Max Act: 3.0454. Min Act: -0.0000

Data Index: 141024 (C4 (Web Text))

Max Activating Token Index: 139