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

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

Neuron 108 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 1.7097. Min Range: -1.7097

Max Act: 1.7097. Min Act: -0.0000

Data Index: 364254 (C4 (Web Text))

Max Activating Token Index: 772

Text #1

Max Range: 1.7097. Min Range: -1.7097

Max Act: 1.0141. Min Act: -0.0001

Data Index: 920864 (C4 (Web Text))

Max Activating Token Index: 582

Text #2

Max Range: 1.7097. Min Range: -1.7097

Max Act: 1.0316. Min Act: -0.0001

Data Index: 1197921 (C4 (Web Text))

Max Activating Token Index: 1004

Text #3

Max Range: 1.7097. Min Range: -1.7097

Max Act: 0.8951. Min Act: -0.0001

Data Index: 755490 (C4 (Web Text))

Max Activating Token Index: 469

Text #4

Max Range: 1.7097. Min Range: -1.7097

Max Act: 0.8971. Min Act: -0.0000

Data Index: 616017 (C4 (Web Text))

Max Activating Token Index: 403

Text #5

Max Range: 1.7097. Min Range: -1.7097

Max Act: 0.9174. Min Act: -0.0000

Data Index: 929234 (C4 (Web Text))

Max Activating Token Index: 295

Text #6

Max Range: 1.7097. Min Range: -1.7097

Max Act: 0.9026. Min Act: -0.0000

Data Index: 1093525 (C4 (Web Text))

Max Activating Token Index: 695

Text #7

Max Range: 1.7097. Min Range: -1.7097

Max Act: 0.8125. Min Act: -0.0001

Data Index: 1138484 (C4 (Web Text))

Max Activating Token Index: 289

Text #8

Max Range: 1.7097. Min Range: -1.7097

Max Act: 0.8410. Min Act: -0.0000

Data Index: 1145020 (C4 (Web Text))

Max Activating Token Index: 575

Text #9

Max Range: 1.7097. Min Range: -1.7097

Max Act: 0.7895. Min Act: -0.0001

Data Index: 541847 (C4 (Web Text))

Max Activating Token Index: 475

Text #10

Max Range: 1.7097. Min Range: -1.7097

Max Act: 0.7869. Min Act: -0.0000

Data Index: 1097993 (C4 (Web Text))

Max Activating Token Index: 486

Text #11

Max Range: 1.7097. Min Range: -1.7097

Max Act: 0.7883. Min Act: -0.0001

Data Index: 874157 (C4 (Web Text))

Max Activating Token Index: 621

Text #12

Max Range: 1.7097. Min Range: -1.7097

Max Act: 0.8344. Min Act: -0.0000

Data Index: 873070 (C4 (Web Text))

Max Activating Token Index: 830

Text #13

Max Range: 1.7097. Min Range: -1.7097

Max Act: 0.7626. Min Act: -0.0000

Data Index: 248349 (C4 (Web Text))

Max Activating Token Index: 759

Text #14

Max Range: 1.7097. Min Range: -1.7097

Max Act: 0.7897. Min Act: -0.0001

Data Index: 804002 (C4 (Web Text))

Max Activating Token Index: 18

Text #15

Max Range: 1.7097. Min Range: -1.7097

Max Act: 0.7066. Min Act: -0.0000

Data Index: 112267 (C4 (Web Text))

Max Activating Token Index: 1018

Text #16

Max Range: 1.7097. Min Range: -1.7097

Max Act: 0.7550. Min Act: -0.0000

Data Index: 92065 (C4 (Web Text))

Max Activating Token Index: 115

Text #17

Max Range: 1.7097. Min Range: -1.7097

Max Act: 0.7661. Min Act: -0.0000

Data Index: 879108 (C4 (Web Text))

Max Activating Token Index: 278

Text #18

Max Range: 1.7097. Min Range: -1.7097

Max Act: 0.7347. Min Act: -0.0001

Data Index: 1212227 (C4 (Web Text))

Max Activating Token Index: 358

Text #19

Max Range: 1.7097. Min Range: -1.7097

Max Act: 0.7867. Min Act: -0.0001

Data Index: 607495 (C4 (Web Text))

Max Activating Token Index: 834