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

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

Neuron 3398 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.0104. Min Range: -0.0104

Max Act: 0.0102. Min Act: -0.0002

Data Index: 1006712 (C4 (Web Text))

Max Activating Token Index: 576

Text #1

Max Range: 0.0104. Min Range: -0.0104

Max Act: 0.0104. Min Act: -0.0002

Data Index: 67224 (C4 (Web Text))

Max Activating Token Index: 560

Text #2

Max Range: 0.0104. Min Range: -0.0104

Max Act: 0.0103. Min Act: -0.0002

Data Index: 409924 (C4 (Web Text))

Max Activating Token Index: 51

Text #3

Max Range: 0.0104. Min Range: -0.0104

Max Act: 0.0100. Min Act: -0.0003

Data Index: 677255 (C4 (Web Text))

Max Activating Token Index: 540

Text #4

Max Range: 0.0104. Min Range: -0.0104

Max Act: 0.0101. Min Act: -0.0003

Data Index: 58034 (C4 (Web Text))

Max Activating Token Index: 487

Text #5

Max Range: 0.0104. Min Range: -0.0104

Max Act: 0.0096. Min Act: -0.0002

Data Index: 1348870 (C4 (Web Text))

Max Activating Token Index: 592

Text #6

Max Range: 0.0104. Min Range: -0.0104

Max Act: 0.0097. Min Act: -0.0003

Data Index: 267707 (C4 (Web Text))

Max Activating Token Index: 669

Text #7

Max Range: 0.0104. Min Range: -0.0104

Max Act: 0.0091. Min Act: -0.0003

Data Index: 218787 (C4 (Web Text))

Max Activating Token Index: 602

Text #8

Max Range: 0.0104. Min Range: -0.0104

Max Act: 0.0094. Min Act: -0.0003

Data Index: 411055 (C4 (Web Text))

Max Activating Token Index: 538

Text #9

Max Range: 0.0104. Min Range: -0.0104

Max Act: 0.0087. Min Act: -0.0002

Data Index: 44210 (C4 (Web Text))

Max Activating Token Index: 305

Text #10

Max Range: 0.0104. Min Range: -0.0104

Max Act: 0.0090. Min Act: -0.0003

Data Index: 1351354 (C4 (Web Text))

Max Activating Token Index: 836

Text #11

Max Range: 0.0104. Min Range: -0.0104

Max Act: 0.0081. Min Act: -0.0003

Data Index: 1487476 (Python Code)

Max Activating Token Index: 672

Text #12

Max Range: 0.0104. Min Range: -0.0104

Max Act: 0.0087. Min Act: -0.0002

Data Index: 639293 (C4 (Web Text))

Max Activating Token Index: 580

Text #13

Max Range: 0.0104. Min Range: -0.0104

Max Act: 0.0088. Min Act: -0.0003

Data Index: 1645319 (Python Code)

Max Activating Token Index: 74

Text #14

Max Range: 0.0104. Min Range: -0.0104

Max Act: 0.0079. Min Act: -0.0002

Data Index: 528210 (C4 (Web Text))

Max Activating Token Index: 206

Text #15

Max Range: 0.0104. Min Range: -0.0104

Max Act: 0.0085. Min Act: -0.0003

Data Index: 273518 (C4 (Web Text))

Max Activating Token Index: 996

Text #16

Max Range: 0.0104. Min Range: -0.0104

Max Act: 0.0086. Min Act: -0.0002

Data Index: 1042351 (C4 (Web Text))

Max Activating Token Index: 260

Text #17

Max Range: 0.0104. Min Range: -0.0104

Max Act: 0.0091. Min Act: -0.0002

Data Index: 814871 (C4 (Web Text))

Max Activating Token Index: 320

Text #18

Max Range: 0.0104. Min Range: -0.0104

Max Act: 0.0086. Min Act: -0.0003

Data Index: 757046 (C4 (Web Text))

Max Activating Token Index: 578

Text #19

Max Range: 0.0104. Min Range: -0.0104

Max Act: 0.0085. Min Act: -0.0003

Data Index: 147234 (C4 (Web Text))

Max Activating Token Index: 217