Model: SoLU Model: 10 Layers, 5120 Neurons per Layer

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

Neuron 80 in Layer 8

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-10l')

Text #0

Max Range: 0.0125. Min Range: -0.0125

Max Act: 0.0125. Min Act: -0.0000

Data Index: 230835 (C4 (Web Text))

Max Activating Token Index: 45

Text #1

Max Range: 0.0125. Min Range: -0.0125

Max Act: 0.0113. Min Act: -0.0000

Data Index: 1261792 (C4 (Web Text))

Max Activating Token Index: 314

Text #2

Max Range: 0.0125. Min Range: -0.0125

Max Act: 0.0114. Min Act: -0.0000

Data Index: 532111 (C4 (Web Text))

Max Activating Token Index: 34

Text #3

Max Range: 0.0125. Min Range: -0.0125

Max Act: 0.0104. Min Act: 0.0000

Data Index: 1133905 (C4 (Web Text))

Max Activating Token Index: 280

Text #4

Max Range: 0.0125. Min Range: -0.0125

Max Act: 0.0107. Min Act: 0.0000

Data Index: 891464 (C4 (Web Text))

Max Activating Token Index: 431

Text #5

Max Range: 0.0125. Min Range: -0.0125

Max Act: 0.0105. Min Act: -0.0000

Data Index: 186734 (C4 (Web Text))

Max Activating Token Index: 870

Text #6

Max Range: 0.0125. Min Range: -0.0125

Max Act: 0.0101. Min Act: -0.0000

Data Index: 1544548 (Python Code)

Max Activating Token Index: 639

Text #7

Max Range: 0.0125. Min Range: -0.0125

Max Act: 0.0110. Min Act: -0.0000

Data Index: 136323 (C4 (Web Text))

Max Activating Token Index: 706

Text #8

Max Range: 0.0125. Min Range: -0.0125

Max Act: 0.0104. Min Act: -0.0000

Data Index: 624177 (C4 (Web Text))

Max Activating Token Index: 664

Text #9

Max Range: 0.0125. Min Range: -0.0125

Max Act: 0.0111. Min Act: -0.0000

Data Index: 577563 (C4 (Web Text))

Max Activating Token Index: 624

Text #10

Max Range: 0.0125. Min Range: -0.0125

Max Act: 0.0099. Min Act: -0.0000

Data Index: 672937 (C4 (Web Text))

Max Activating Token Index: 791

Text #11

Max Range: 0.0125. Min Range: -0.0125

Max Act: 0.0102. Min Act: 0.0000

Data Index: 925481 (C4 (Web Text))

Max Activating Token Index: 12

Text #12

Max Range: 0.0125. Min Range: -0.0125

Max Act: 0.0102. Min Act: -0.0000

Data Index: 941934 (C4 (Web Text))

Max Activating Token Index: 889

Text #13

Max Range: 0.0125. Min Range: -0.0125

Max Act: 0.0100. Min Act: -0.0000

Data Index: 848185 (C4 (Web Text))

Max Activating Token Index: 450

Text #14

Max Range: 0.0125. Min Range: -0.0125

Max Act: 0.0103. Min Act: 0.0001

Data Index: 1200919 (C4 (Web Text))

Max Activating Token Index: 299

Text #15

Max Range: 0.0125. Min Range: -0.0125

Max Act: 0.0104. Min Act: -0.0000

Data Index: 1425992 (Python Code)

Max Activating Token Index: 697

Text #16

Max Range: 0.0125. Min Range: -0.0125

Max Act: 0.0098. Min Act: 0.0001

Data Index: 131187 (C4 (Web Text))

Max Activating Token Index: 18

Text #17

Max Range: 0.0125. Min Range: -0.0125

Max Act: 0.0102. Min Act: -0.0000

Data Index: 1359093 (C4 (Web Text))

Max Activating Token Index: 826

Text #18

Max Range: 0.0125. Min Range: -0.0125

Max Act: 0.0096. Min Act: -0.0000

Data Index: 952923 (C4 (Web Text))

Max Activating Token Index: 721

Text #19

Max Range: 0.0125. Min Range: -0.0125

Max Act: 0.0097. Min Act: -0.0000

Data Index: 507147 (C4 (Web Text))

Max Activating Token Index: 641