Model: SoLU Model: 3 Layers, 2048 Neurons per Layer

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

Neuron 423 in Layer 0

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-3l')

Text #0

Max Range: 0.0061. Min Range: -0.0061

Max Act: 0.0061. Min Act: -0.0006

Data Index: 963725 (C4 (Web Text))

Max Activating Token Index: 772

Text #1

Max Range: 0.0061. Min Range: -0.0061

Max Act: 0.0060. Min Act: -0.0006

Data Index: 1445124 (Python Code)

Max Activating Token Index: 40

Text #2

Max Range: 0.0061. Min Range: -0.0061

Max Act: 0.0058. Min Act: -0.0006

Data Index: 1592155 (Python Code)

Max Activating Token Index: 253

Text #3

Max Range: 0.0061. Min Range: -0.0061

Max Act: 0.0059. Min Act: -0.0005

Data Index: 1265692 (C4 (Web Text))

Max Activating Token Index: 101

Text #4

Max Range: 0.0061. Min Range: -0.0061

Max Act: 0.0058. Min Act: -0.0006

Data Index: 1381301 (Python Code)

Max Activating Token Index: 364

Text #5

Max Range: 0.0061. Min Range: -0.0061

Max Act: 0.0057. Min Act: -0.0006

Data Index: 21199 (C4 (Web Text))

Max Activating Token Index: 496

Text #6

Max Range: 0.0061. Min Range: -0.0061

Max Act: 0.0057. Min Act: -0.0006

Data Index: 1242321 (C4 (Web Text))

Max Activating Token Index: 599

Text #7

Max Range: 0.0061. Min Range: -0.0061

Max Act: 0.0057. Min Act: -0.0005

Data Index: 1645769 (Python Code)

Max Activating Token Index: 1023

Text #8

Max Range: 0.0061. Min Range: -0.0061

Max Act: 0.0057. Min Act: -0.0006

Data Index: 1258591 (C4 (Web Text))

Max Activating Token Index: 374

Text #9

Max Range: 0.0061. Min Range: -0.0061

Max Act: 0.0057. Min Act: -0.0006

Data Index: 202296 (C4 (Web Text))

Max Activating Token Index: 112

Text #10

Max Range: 0.0061. Min Range: -0.0061

Max Act: 0.0055. Min Act: -0.0006

Data Index: 1389936 (Python Code)

Max Activating Token Index: 642

Text #11

Max Range: 0.0061. Min Range: -0.0061

Max Act: 0.0056. Min Act: -0.0006

Data Index: 1290484 (C4 (Web Text))

Max Activating Token Index: 336

Text #12

Max Range: 0.0061. Min Range: -0.0061

Max Act: 0.0057. Min Act: -0.0006

Data Index: 1401990 (Python Code)

Max Activating Token Index: 943

Text #13

Max Range: 0.0061. Min Range: -0.0061

Max Act: 0.0056. Min Act: -0.0005

Data Index: 98338 (C4 (Web Text))

Max Activating Token Index: 892

Text #14

Max Range: 0.0061. Min Range: -0.0061

Max Act: 0.0055. Min Act: -0.0006

Data Index: 1226288 (C4 (Web Text))

Max Activating Token Index: 611

Text #15

Max Range: 0.0061. Min Range: -0.0061

Max Act: 0.0056. Min Act: -0.0006

Data Index: 286027 (C4 (Web Text))

Max Activating Token Index: 650

Text #16

Max Range: 0.0061. Min Range: -0.0061

Max Act: 0.0055. Min Act: -0.0006

Data Index: 807555 (C4 (Web Text))

Max Activating Token Index: 1023

Text #17

Max Range: 0.0061. Min Range: -0.0061

Max Act: 0.0055. Min Act: -0.0006

Data Index: 1362580 (Python Code)

Max Activating Token Index: 180

Text #18

Max Range: 0.0061. Min Range: -0.0061

Max Act: 0.0055. Min Act: -0.0006

Data Index: 888110 (C4 (Web Text))

Max Activating Token Index: 200

Text #19

Max Range: 0.0061. Min Range: -0.0061

Max Act: 0.0055. Min Act: -0.0005

Data Index: 1585178 (Python Code)

Max Activating Token Index: 1023