Model: SoLU Model: 12 Layers, 6144 Neurons per Layer

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

Neuron 5661 in Layer 9

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-12l')

Text #0

Max Range: 4.7233. Min Range: -4.7233

Max Act: 4.6539. Min Act: -0.0000

Data Index: 879478 (C4 (Web Text))

Max Activating Token Index: 892

Text #1

Max Range: 4.7233. Min Range: -4.7233

Max Act: 4.7233. Min Act: -0.0000

Data Index: 920434 (C4 (Web Text))

Max Activating Token Index: 786

Text #2

Max Range: 4.7233. Min Range: -4.7233

Max Act: 4.4287. Min Act: -0.0000

Data Index: 1448697 (Python Code)

Max Activating Token Index: 167

Text #3

Max Range: 4.7233. Min Range: -4.7233

Max Act: 4.2585. Min Act: -0.0000

Data Index: 1542421 (Python Code)

Max Activating Token Index: 171

Text #4

Max Range: 4.7233. Min Range: -4.7233

Max Act: 4.2180. Min Act: -0.0000

Data Index: 672358 (C4 (Web Text))

Max Activating Token Index: 713

Text #5

Max Range: 4.7233. Min Range: -4.7233

Max Act: 4.2047. Min Act: -0.0000

Data Index: 1412801 (Python Code)

Max Activating Token Index: 208

Text #6

Max Range: 4.7233. Min Range: -4.7233

Max Act: 3.9051. Min Act: -0.0000

Data Index: 148334 (C4 (Web Text))

Max Activating Token Index: 631

Text #7

Max Range: 4.7233. Min Range: -4.7233

Max Act: 4.0650. Min Act: -0.0000

Data Index: 178593 (C4 (Web Text))

Max Activating Token Index: 632

Text #8

Max Range: 4.7233. Min Range: -4.7233

Max Act: 4.0080. Min Act: -0.0000

Data Index: 1517582 (Python Code)

Max Activating Token Index: 94

Text #9

Max Range: 4.7233. Min Range: -4.7233

Max Act: 4.1477. Min Act: -0.0000

Data Index: 1313003 (C4 (Web Text))

Max Activating Token Index: 659

Text #10

Max Range: 4.7233. Min Range: -4.7233

Max Act: 3.7875. Min Act: -0.0000

Data Index: 1301783 (C4 (Web Text))

Max Activating Token Index: 113

Text #11

Max Range: 4.7233. Min Range: -4.7233

Max Act: 3.7892. Min Act: -0.0000

Data Index: 1364761 (Python Code)

Max Activating Token Index: 96

Text #12

Max Range: 4.7233. Min Range: -4.7233

Max Act: 3.7370. Min Act: -0.0000

Data Index: 1306838 (C4 (Web Text))

Max Activating Token Index: 87

Text #13

Max Range: 4.7233. Min Range: -4.7233

Max Act: 3.8874. Min Act: -0.0000

Data Index: 1518929 (Python Code)

Max Activating Token Index: 167

Text #14

Max Range: 4.7233. Min Range: -4.7233

Max Act: 4.0976. Min Act: -0.0000

Data Index: 1447075 (Python Code)

Max Activating Token Index: 236

Text #15

Max Range: 4.7233. Min Range: -4.7233

Max Act: 3.7549. Min Act: -0.0000

Data Index: 1295591 (C4 (Web Text))

Max Activating Token Index: 743

Text #16

Max Range: 4.7233. Min Range: -4.7233

Max Act: 3.8337. Min Act: -0.0000

Data Index: 677412 (C4 (Web Text))

Max Activating Token Index: 1013

Text #17

Max Range: 4.7233. Min Range: -4.7233

Max Act: 3.8222. Min Act: -0.0000

Data Index: 1439803 (Python Code)

Max Activating Token Index: 826

Text #18

Max Range: 4.7233. Min Range: -4.7233

Max Act: 3.6453. Min Act: -0.0000

Data Index: 447114 (C4 (Web Text))

Max Activating Token Index: 743

Text #19

Max Range: 4.7233. Min Range: -4.7233

Max Act: 3.8387. Min Act: -0.0000

Data Index: 1583334 (Python Code)

Max Activating Token Index: 187