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

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

Neuron 2536 in Layer 6

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.1263. Min Range: -0.1263

Max Act: 0.1210. Min Act: -0.0001

Data Index: 483282 (C4 (Web Text))

Max Activating Token Index: 413

Text #1

Max Range: 0.1263. Min Range: -0.1263

Max Act: 0.1041. Min Act: -0.0001

Data Index: 1560697 (Python Code)

Max Activating Token Index: 396

Text #2

Max Range: 0.1263. Min Range: -0.1263

Max Act: 0.1263. Min Act: -0.0001

Data Index: 484734 (C4 (Web Text))

Max Activating Token Index: 306

Text #3

Max Range: 0.1263. Min Range: -0.1263

Max Act: 0.1088. Min Act: -0.0001

Data Index: 160982 (C4 (Web Text))

Max Activating Token Index: 838

Text #4

Max Range: 0.1263. Min Range: -0.1263

Max Act: 0.0937. Min Act: -0.0001

Data Index: 524608 (C4 (Web Text))

Max Activating Token Index: 732

Text #5

Max Range: 0.1263. Min Range: -0.1263

Max Act: 0.1026. Min Act: -0.0000

Data Index: 10095 (C4 (Web Text))

Max Activating Token Index: 160

Text #6

Max Range: 0.1263. Min Range: -0.1263

Max Act: 0.0974. Min Act: -0.0000

Data Index: 160004 (C4 (Web Text))

Max Activating Token Index: 125

Text #7

Max Range: 0.1263. Min Range: -0.1263

Max Act: 0.0920. Min Act: -0.0000

Data Index: 1275809 (C4 (Web Text))

Max Activating Token Index: 950

Text #8

Max Range: 0.1263. Min Range: -0.1263

Max Act: 0.0968. Min Act: -0.0001

Data Index: 1028016 (C4 (Web Text))

Max Activating Token Index: 842

Text #9

Max Range: 0.1263. Min Range: -0.1263

Max Act: 0.0952. Min Act: -0.0001

Data Index: 1420864 (Python Code)

Max Activating Token Index: 844

Text #10

Max Range: 0.1263. Min Range: -0.1263

Max Act: 0.0960. Min Act: -0.0001

Data Index: 1486609 (Python Code)

Max Activating Token Index: 361

Text #11

Max Range: 0.1263. Min Range: -0.1263

Max Act: 0.0939. Min Act: -0.0001

Data Index: 347081 (C4 (Web Text))

Max Activating Token Index: 514

Text #12

Max Range: 0.1263. Min Range: -0.1263

Max Act: 0.0831. Min Act: -0.0001

Data Index: 186921 (C4 (Web Text))

Max Activating Token Index: 295

Text #13

Max Range: 0.1263. Min Range: -0.1263

Max Act: 0.0838. Min Act: -0.0001

Data Index: 1382487 (Python Code)

Max Activating Token Index: 740

Text #14

Max Range: 0.1263. Min Range: -0.1263

Max Act: 0.0822. Min Act: -0.0001

Data Index: 224920 (C4 (Web Text))

Max Activating Token Index: 816

Text #15

Max Range: 0.1263. Min Range: -0.1263

Max Act: 0.0815. Min Act: -0.0001

Data Index: 1430920 (Python Code)

Max Activating Token Index: 233

Text #16

Max Range: 0.1263. Min Range: -0.1263

Max Act: 0.0809. Min Act: -0.0001

Data Index: 968101 (C4 (Web Text))

Max Activating Token Index: 121

Text #17

Max Range: 0.1263. Min Range: -0.1263

Max Act: 0.0832. Min Act: -0.0000

Data Index: 1565504 (Python Code)

Max Activating Token Index: 927

Text #18

Max Range: 0.1263. Min Range: -0.1263

Max Act: 0.0762. Min Act: -0.0001

Data Index: 1103170 (C4 (Web Text))

Max Activating Token Index: 151

Text #19

Max Range: 0.1263. Min Range: -0.1263

Max Act: 0.0829. Min Act: -0.0001

Data Index: 1098710 (C4 (Web Text))

Max Activating Token Index: 560