Model: SoLU Model: 2 Layers, 2048 Neurons per Layer

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

Neuron 850 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-2l')

Text #0

Max Range: 0.0107. Min Range: -0.0107

Max Act: 0.0105. Min Act: -0.0003

Data Index: 1603407 (Python Code)

Max Activating Token Index: 904

Text #1

Max Range: 0.0107. Min Range: -0.0107

Max Act: 0.0107. Min Act: -0.0003

Data Index: 1579553 (Python Code)

Max Activating Token Index: 280

Text #2

Max Range: 0.0107. Min Range: -0.0107

Max Act: 0.0094. Min Act: -0.0003

Data Index: 1360507 (Python Code)

Max Activating Token Index: 943

Text #3

Max Range: 0.0107. Min Range: -0.0107

Max Act: 0.0093. Min Act: -0.0001

Data Index: 1563855 (Python Code)

Max Activating Token Index: 541

Text #4

Max Range: 0.0107. Min Range: -0.0107

Max Act: 0.0092. Min Act: -0.0003

Data Index: 823777 (C4 (Web Text))

Max Activating Token Index: 211

Text #5

Max Range: 0.0107. Min Range: -0.0107

Max Act: 0.0090. Min Act: -0.0003

Data Index: 1070235 (C4 (Web Text))

Max Activating Token Index: 930

Text #6

Max Range: 0.0107. Min Range: -0.0107

Max Act: 0.0088. Min Act: -0.0003

Data Index: 1646180 (Python Code)

Max Activating Token Index: 648

Text #7

Max Range: 0.0107. Min Range: -0.0107

Max Act: 0.0090. Min Act: -0.0002

Data Index: 1181434 (C4 (Web Text))

Max Activating Token Index: 580

Text #8

Max Range: 0.0107. Min Range: -0.0107

Max Act: 0.0089. Min Act: -0.0003

Data Index: 1398680 (Python Code)

Max Activating Token Index: 424

Text #9

Max Range: 0.0107. Min Range: -0.0107

Max Act: 0.0088. Min Act: -0.0003

Data Index: 751239 (C4 (Web Text))

Max Activating Token Index: 559

Text #10

Max Range: 0.0107. Min Range: -0.0107

Max Act: 0.0083. Min Act: -0.0003

Data Index: 1511508 (Python Code)

Max Activating Token Index: 997

Text #11

Max Range: 0.0107. Min Range: -0.0107

Max Act: 0.0080. Min Act: -0.0003

Data Index: 1426641 (Python Code)

Max Activating Token Index: 846

Text #12

Max Range: 0.0107. Min Range: -0.0107

Max Act: 0.0082. Min Act: -0.0002

Data Index: 89716 (C4 (Web Text))

Max Activating Token Index: 974

Text #13

Max Range: 0.0107. Min Range: -0.0107

Max Act: 0.0081. Min Act: -0.0003

Data Index: 1336308 (C4 (Web Text))

Max Activating Token Index: 647

Text #14

Max Range: 0.0107. Min Range: -0.0107

Max Act: 0.0082. Min Act: -0.0002

Data Index: 1120189 (C4 (Web Text))

Max Activating Token Index: 158

Text #15

Max Range: 0.0107. Min Range: -0.0107

Max Act: 0.0080. Min Act: -0.0003

Data Index: 1630680 (Python Code)

Max Activating Token Index: 980

Text #16

Max Range: 0.0107. Min Range: -0.0107

Max Act: 0.0082. Min Act: -0.0003

Data Index: 700086 (C4 (Web Text))

Max Activating Token Index: 895

Text #17

Max Range: 0.0107. Min Range: -0.0107

Max Act: 0.0082. Min Act: -0.0002

Data Index: 684911 (C4 (Web Text))

Max Activating Token Index: 350

Text #18

Max Range: 0.0107. Min Range: -0.0107

Max Act: 0.0081. Min Act: -0.0002

Data Index: 1239848 (C4 (Web Text))

Max Activating Token Index: 274

Text #19

Max Range: 0.0107. Min Range: -0.0107

Max Act: 0.0081. Min Act: -0.0003

Data Index: 1527059 (Python Code)

Max Activating Token Index: 243