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

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

Neuron 5111 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: 0.0099. Min Range: -0.0099

Max Act: 0.0099. Min Act: -0.0003

Data Index: 879049 (C4 (Web Text))

Max Activating Token Index: 161

Text #1

Max Range: 0.0099. Min Range: -0.0099

Max Act: 0.0097. Min Act: -0.0003

Data Index: 54659 (C4 (Web Text))

Max Activating Token Index: 981

Text #2

Max Range: 0.0099. Min Range: -0.0099

Max Act: 0.0098. Min Act: -0.0003

Data Index: 18578 (C4 (Web Text))

Max Activating Token Index: 443

Text #3

Max Range: 0.0099. Min Range: -0.0099

Max Act: 0.0094. Min Act: -0.0003

Data Index: 324780 (C4 (Web Text))

Max Activating Token Index: 261

Text #4

Max Range: 0.0099. Min Range: -0.0099

Max Act: 0.0096. Min Act: -0.0003

Data Index: 856898 (C4 (Web Text))

Max Activating Token Index: 616

Text #5

Max Range: 0.0099. Min Range: -0.0099

Max Act: 0.0094. Min Act: -0.0003

Data Index: 1385929 (Python Code)

Max Activating Token Index: 208

Text #6

Max Range: 0.0099. Min Range: -0.0099

Max Act: 0.0090. Min Act: -0.0003

Data Index: 960827 (C4 (Web Text))

Max Activating Token Index: 29

Text #7

Max Range: 0.0099. Min Range: -0.0099

Max Act: 0.0093. Min Act: -0.0002

Data Index: 1602712 (Python Code)

Max Activating Token Index: 269

Text #8

Max Range: 0.0099. Min Range: -0.0099

Max Act: 0.0090. Min Act: -0.0003

Data Index: 1218267 (C4 (Web Text))

Max Activating Token Index: 870

Text #9

Max Range: 0.0099. Min Range: -0.0099

Max Act: 0.0095. Min Act: -0.0003

Data Index: 1540656 (Python Code)

Max Activating Token Index: 802

Text #10

Max Range: 0.0099. Min Range: -0.0099

Max Act: 0.0091. Min Act: -0.0003

Data Index: 1249463 (C4 (Web Text))

Max Activating Token Index: 55

Text #11

Max Range: 0.0099. Min Range: -0.0099

Max Act: 0.0091. Min Act: -0.0003

Data Index: 414797 (C4 (Web Text))

Max Activating Token Index: 620

Text #12

Max Range: 0.0099. Min Range: -0.0099

Max Act: 0.0088. Min Act: -0.0003

Data Index: 1149726 (C4 (Web Text))

Max Activating Token Index: 570

Text #13

Max Range: 0.0099. Min Range: -0.0099

Max Act: 0.0090. Min Act: -0.0003

Data Index: 794178 (C4 (Web Text))

Max Activating Token Index: 639

Text #14

Max Range: 0.0099. Min Range: -0.0099

Max Act: 0.0089. Min Act: -0.0003

Data Index: 1340936 (C4 (Web Text))

Max Activating Token Index: 516

Text #15

Max Range: 0.0099. Min Range: -0.0099

Max Act: 0.0088. Min Act: -0.0003

Data Index: 1234303 (C4 (Web Text))

Max Activating Token Index: 524

Text #16

Max Range: 0.0099. Min Range: -0.0099

Max Act: 0.0090. Min Act: -0.0003

Data Index: 239162 (C4 (Web Text))

Max Activating Token Index: 274

Text #17

Max Range: 0.0099. Min Range: -0.0099

Max Act: 0.0089. Min Act: -0.0003

Data Index: 1434579 (Python Code)

Max Activating Token Index: 63

Text #18

Max Range: 0.0099. Min Range: -0.0099

Max Act: 0.0086. Min Act: -0.0003

Data Index: 977491 (C4 (Web Text))

Max Activating Token Index: 811

Text #19

Max Range: 0.0099. Min Range: -0.0099

Max Act: 0.0086. Min Act: -0.0003

Data Index: 614084 (C4 (Web Text))

Max Activating Token Index: 515