Model: SoLU Model: 4 Layers, 2048 Neurons per Layer

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

Neuron 350 in Layer 0

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-4l')

Text #0

Max Range: 0.0111. Min Range: -0.0111

Max Act: 0.0111. Min Act: -0.0005

Data Index: 1001986 (C4 (Web Text))

Max Activating Token Index: 711

Text #1

Max Range: 0.0111. Min Range: -0.0111

Max Act: 0.0105. Min Act: -0.0005

Data Index: 102490 (C4 (Web Text))

Max Activating Token Index: 139

Text #2

Max Range: 0.0111. Min Range: -0.0111

Max Act: 0.0102. Min Act: -0.0005

Data Index: 704145 (C4 (Web Text))

Max Activating Token Index: 470

Text #3

Max Range: 0.0111. Min Range: -0.0111

Max Act: 0.0095. Min Act: -0.0006

Data Index: 744193 (C4 (Web Text))

Max Activating Token Index: 394

Text #4

Max Range: 0.0111. Min Range: -0.0111

Max Act: 0.0094. Min Act: -0.0005

Data Index: 444328 (C4 (Web Text))

Max Activating Token Index: 607

Text #5

Max Range: 0.0111. Min Range: -0.0111

Max Act: 0.0094. Min Act: -0.0005

Data Index: 898636 (C4 (Web Text))

Max Activating Token Index: 218

Text #6

Max Range: 0.0111. Min Range: -0.0111

Max Act: 0.0093. Min Act: -0.0005

Data Index: 42740 (C4 (Web Text))

Max Activating Token Index: 263

Text #7

Max Range: 0.0111. Min Range: -0.0111

Max Act: 0.0091. Min Act: -0.0005

Data Index: 817969 (C4 (Web Text))

Max Activating Token Index: 845

Text #8

Max Range: 0.0111. Min Range: -0.0111

Max Act: 0.0090. Min Act: -0.0005

Data Index: 728004 (C4 (Web Text))

Max Activating Token Index: 231

Text #9

Max Range: 0.0111. Min Range: -0.0111

Max Act: 0.0092. Min Act: -0.0005

Data Index: 162150 (C4 (Web Text))

Max Activating Token Index: 782

Text #10

Max Range: 0.0111. Min Range: -0.0111

Max Act: 0.0092. Min Act: -0.0005

Data Index: 407735 (C4 (Web Text))

Max Activating Token Index: 52

Text #11

Max Range: 0.0111. Min Range: -0.0111

Max Act: 0.0091. Min Act: -0.0005

Data Index: 396557 (C4 (Web Text))

Max Activating Token Index: 365

Text #12

Max Range: 0.0111. Min Range: -0.0111

Max Act: 0.0090. Min Act: -0.0005

Data Index: 950887 (C4 (Web Text))

Max Activating Token Index: 599

Text #13

Max Range: 0.0111. Min Range: -0.0111

Max Act: 0.0091. Min Act: -0.0005

Data Index: 109876 (C4 (Web Text))

Max Activating Token Index: 53

Text #14

Max Range: 0.0111. Min Range: -0.0111

Max Act: 0.0090. Min Act: -0.0005

Data Index: 1590108 (Python Code)

Max Activating Token Index: 236

Text #15

Max Range: 0.0111. Min Range: -0.0111

Max Act: 0.0089. Min Act: -0.0006

Data Index: 243050 (C4 (Web Text))

Max Activating Token Index: 222

Text #16

Max Range: 0.0111. Min Range: -0.0111

Max Act: 0.0090. Min Act: -0.0005

Data Index: 1127160 (C4 (Web Text))

Max Activating Token Index: 447

Text #17

Max Range: 0.0111. Min Range: -0.0111

Max Act: 0.0089. Min Act: -0.0005

Data Index: 456616 (C4 (Web Text))

Max Activating Token Index: 896

Text #18

Max Range: 0.0111. Min Range: -0.0111

Max Act: 0.0090. Min Act: -0.0005

Data Index: 362926 (C4 (Web Text))

Max Activating Token Index: 507

Text #19

Max Range: 0.0111. Min Range: -0.0111

Max Act: 0.0091. Min Act: -0.0005

Data Index: 1267978 (C4 (Web Text))

Max Activating Token Index: 690