Model: SoLU Model: 8 Layers, 4096 Neurons per Layer

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

Neuron 3324 in Layer 7

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-8l')

Text #0

Max Range: 0.4245. Min Range: -0.4245

Max Act: 0.4245. Min Act: -0.0001

Data Index: 896373 (C4 (Web Text))

Max Activating Token Index: 62

Text #1

Max Range: 0.4245. Min Range: -0.4245

Max Act: 0.3410. Min Act: -0.0001

Data Index: 1603995 (Python Code)

Max Activating Token Index: 103

Text #2

Max Range: 0.4245. Min Range: -0.4245

Max Act: 0.3059. Min Act: -0.0001

Data Index: 725659 (C4 (Web Text))

Max Activating Token Index: 638

Text #3

Max Range: 0.4245. Min Range: -0.4245

Max Act: 0.3644. Min Act: -0.0001

Data Index: 746063 (C4 (Web Text))

Max Activating Token Index: 176

Text #4

Max Range: 0.4245. Min Range: -0.4245

Max Act: 0.3103. Min Act: -0.0001

Data Index: 1582644 (Python Code)

Max Activating Token Index: 228

Text #5

Max Range: 0.4245. Min Range: -0.4245

Max Act: 0.3199. Min Act: -0.0001

Data Index: 1483044 (Python Code)

Max Activating Token Index: 537

Text #6

Max Range: 0.4245. Min Range: -0.4245

Max Act: 0.3071. Min Act: -0.0001

Data Index: 1564397 (Python Code)

Max Activating Token Index: 95

Text #7

Max Range: 0.4245. Min Range: -0.4245

Max Act: 0.3225. Min Act: -0.0001

Data Index: 1210573 (C4 (Web Text))

Max Activating Token Index: 387

Text #8

Max Range: 0.4245. Min Range: -0.4245

Max Act: 0.3052. Min Act: -0.0001

Data Index: 1475001 (Python Code)

Max Activating Token Index: 214

Text #9

Max Range: 0.4245. Min Range: -0.4245

Max Act: 0.3089. Min Act: -0.0001

Data Index: 105862 (C4 (Web Text))

Max Activating Token Index: 1008

Text #10

Max Range: 0.4245. Min Range: -0.4245

Max Act: 0.3021. Min Act: -0.0001

Data Index: 102959 (C4 (Web Text))

Max Activating Token Index: 140

Text #11

Max Range: 0.4245. Min Range: -0.4245

Max Act: 0.2960. Min Act: -0.0001

Data Index: 793231 (C4 (Web Text))

Max Activating Token Index: 127

Text #12

Max Range: 0.4245. Min Range: -0.4245

Max Act: 0.2959. Min Act: -0.0001

Data Index: 69851 (C4 (Web Text))

Max Activating Token Index: 32

Text #13

Max Range: 0.4245. Min Range: -0.4245

Max Act: 0.2852. Min Act: -0.0001

Data Index: 789694 (C4 (Web Text))

Max Activating Token Index: 310

Text #14

Max Range: 0.4245. Min Range: -0.4245

Max Act: 0.2946. Min Act: -0.0001

Data Index: 1622116 (Python Code)

Max Activating Token Index: 607

Text #15

Max Range: 0.4245. Min Range: -0.4245

Max Act: 0.2906. Min Act: -0.0001

Data Index: 1265052 (C4 (Web Text))

Max Activating Token Index: 110

Text #16

Max Range: 0.4245. Min Range: -0.4245

Max Act: 0.3029. Min Act: -0.0001

Data Index: 133084 (C4 (Web Text))

Max Activating Token Index: 469

Text #17

Max Range: 0.4245. Min Range: -0.4245

Max Act: 0.2786. Min Act: -0.0001

Data Index: 1482844 (Python Code)

Max Activating Token Index: 175

Text #18

Max Range: 0.4245. Min Range: -0.4245

Max Act: 0.2870. Min Act: -0.0001

Data Index: 1554331 (Python Code)

Max Activating Token Index: 11

Text #19

Max Range: 0.4245. Min Range: -0.4245

Max Act: 0.2961. Min Act: -0.0001

Data Index: 1423837 (Python Code)

Max Activating Token Index: 49