Model: SoLU Model: 8 Layers, 4096 Neurons per Layer

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

Neuron 3092 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-8l')

Text #0

Max Range: 0.5701. Min Range: -0.5701

Max Act: 0.5701. Min Act: -0.0001

Data Index: 75341 (C4 (Web Text))

Max Activating Token Index: 624

Text #1

Max Range: 0.5701. Min Range: -0.5701

Max Act: 0.5056. Min Act: -0.0001

Data Index: 987536 (C4 (Web Text))

Max Activating Token Index: 197

Text #2

Max Range: 0.5701. Min Range: -0.5701

Max Act: 0.4568. Min Act: -0.0001

Data Index: 885111 (C4 (Web Text))

Max Activating Token Index: 973

Text #3

Max Range: 0.5701. Min Range: -0.5701

Max Act: 0.4873. Min Act: -0.0001

Data Index: 500026 (C4 (Web Text))

Max Activating Token Index: 783

Text #4

Max Range: 0.5701. Min Range: -0.5701

Max Act: 0.4352. Min Act: -0.0001

Data Index: 422017 (C4 (Web Text))

Max Activating Token Index: 971

Text #5

Max Range: 0.5701. Min Range: -0.5701

Max Act: 0.4365. Min Act: -0.0001

Data Index: 296324 (C4 (Web Text))

Max Activating Token Index: 88

Text #6

Max Range: 0.5701. Min Range: -0.5701

Max Act: 0.4196. Min Act: -0.0001

Data Index: 1552599 (Python Code)

Max Activating Token Index: 314

Text #7

Max Range: 0.5701. Min Range: -0.5701

Max Act: 0.4167. Min Act: -0.0001

Data Index: 921193 (C4 (Web Text))

Max Activating Token Index: 835

Text #8

Max Range: 0.5701. Min Range: -0.5701

Max Act: 0.4150. Min Act: -0.0001

Data Index: 816808 (C4 (Web Text))

Max Activating Token Index: 626

Text #9

Max Range: 0.5701. Min Range: -0.5701

Max Act: 0.3955. Min Act: -0.0001

Data Index: 333837 (C4 (Web Text))

Max Activating Token Index: 430

Text #10

Max Range: 0.5701. Min Range: -0.5701

Max Act: 0.4106. Min Act: -0.0001

Data Index: 1178929 (C4 (Web Text))

Max Activating Token Index: 468

Text #11

Max Range: 0.5701. Min Range: -0.5701

Max Act: 0.3844. Min Act: -0.0001

Data Index: 170859 (C4 (Web Text))

Max Activating Token Index: 651

Text #12

Max Range: 0.5701. Min Range: -0.5701

Max Act: 0.3809. Min Act: -0.0001

Data Index: 394085 (C4 (Web Text))

Max Activating Token Index: 209

Text #13

Max Range: 0.5701. Min Range: -0.5701

Max Act: 0.3892. Min Act: -0.0001

Data Index: 34352 (C4 (Web Text))

Max Activating Token Index: 310

Text #14

Max Range: 0.5701. Min Range: -0.5701

Max Act: 0.3642. Min Act: -0.0001

Data Index: 1501035 (Python Code)

Max Activating Token Index: 610

Text #15

Max Range: 0.5701. Min Range: -0.5701

Max Act: 0.3771. Min Act: -0.0001

Data Index: 891321 (C4 (Web Text))

Max Activating Token Index: 754

Text #16

Max Range: 0.5701. Min Range: -0.5701

Max Act: 0.3766. Min Act: -0.0001

Data Index: 1179695 (C4 (Web Text))

Max Activating Token Index: 939

Text #17

Max Range: 0.5701. Min Range: -0.5701

Max Act: 0.3760. Min Act: -0.0001

Data Index: 1098164 (C4 (Web Text))

Max Activating Token Index: 169

Text #18

Max Range: 0.5701. Min Range: -0.5701

Max Act: 0.3650. Min Act: -0.0001

Data Index: 763454 (C4 (Web Text))

Max Activating Token Index: 895

Text #19

Max Range: 0.5701. Min Range: -0.5701

Max Act: 0.3620. Min Act: -0.0001

Data Index: 838630 (C4 (Web Text))

Max Activating Token Index: 126