Model: SoLU Model: 4 Layers, 2048 Neurons per Layer

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

Neuron 426 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-4l')

Text #0

Max Range: 0.0055. Min Range: -0.0055

Max Act: 0.0053. Min Act: -0.0002

Data Index: 1452359 (Python Code)

Max Activating Token Index: 295

Text #1

Max Range: 0.0055. Min Range: -0.0055

Max Act: 0.0055. Min Act: -0.0002

Data Index: 841147 (C4 (Web Text))

Max Activating Token Index: 473

Text #2

Max Range: 0.0055. Min Range: -0.0055

Max Act: 0.0054. Min Act: -0.0002

Data Index: 87234 (C4 (Web Text))

Max Activating Token Index: 622

Text #3

Max Range: 0.0055. Min Range: -0.0055

Max Act: 0.0050. Min Act: -0.0002

Data Index: 838965 (C4 (Web Text))

Max Activating Token Index: 871

Text #4

Max Range: 0.0055. Min Range: -0.0055

Max Act: 0.0048. Min Act: -0.0002

Data Index: 323818 (C4 (Web Text))

Max Activating Token Index: 131

Text #5

Max Range: 0.0055. Min Range: -0.0055

Max Act: 0.0054. Min Act: -0.0002

Data Index: 905198 (C4 (Web Text))

Max Activating Token Index: 362

Text #6

Max Range: 0.0055. Min Range: -0.0055

Max Act: 0.0054. Min Act: -0.0002

Data Index: 159320 (C4 (Web Text))

Max Activating Token Index: 500

Text #7

Max Range: 0.0055. Min Range: -0.0055

Max Act: 0.0043. Min Act: -0.0002

Data Index: 800111 (C4 (Web Text))

Max Activating Token Index: 378

Text #8

Max Range: 0.0055. Min Range: -0.0055

Max Act: 0.0042. Min Act: -0.0002

Data Index: 985609 (C4 (Web Text))

Max Activating Token Index: 652

Text #9

Max Range: 0.0055. Min Range: -0.0055

Max Act: 0.0043. Min Act: -0.0002

Data Index: 353471 (C4 (Web Text))

Max Activating Token Index: 962

Text #10

Max Range: 0.0055. Min Range: -0.0055

Max Act: 0.0038. Min Act: -0.0002

Data Index: 1228680 (C4 (Web Text))

Max Activating Token Index: 757

Text #11

Max Range: 0.0055. Min Range: -0.0055

Max Act: 0.0041. Min Act: -0.0002

Data Index: 668580 (C4 (Web Text))

Max Activating Token Index: 641

Text #12

Max Range: 0.0055. Min Range: -0.0055

Max Act: 0.0037. Min Act: -0.0002

Data Index: 656311 (C4 (Web Text))

Max Activating Token Index: 583

Text #13

Max Range: 0.0055. Min Range: -0.0055

Max Act: 0.0040. Min Act: -0.0002

Data Index: 160047 (C4 (Web Text))

Max Activating Token Index: 559

Text #14

Max Range: 0.0055. Min Range: -0.0055

Max Act: 0.0040. Min Act: -0.0001

Data Index: 172937 (C4 (Web Text))

Max Activating Token Index: 310

Text #15

Max Range: 0.0055. Min Range: -0.0055

Max Act: 0.0033. Min Act: -0.0002

Data Index: 1150415 (C4 (Web Text))

Max Activating Token Index: 217

Text #16

Max Range: 0.0055. Min Range: -0.0055

Max Act: 0.0036. Min Act: -0.0002

Data Index: 1325548 (C4 (Web Text))

Max Activating Token Index: 940

Text #17

Max Range: 0.0055. Min Range: -0.0055

Max Act: 0.0036. Min Act: -0.0002

Data Index: 881809 (C4 (Web Text))

Max Activating Token Index: 431

Text #18

Max Range: 0.0055. Min Range: -0.0055

Max Act: 0.0032. Min Act: -0.0002

Data Index: 520590 (C4 (Web Text))

Max Activating Token Index: 960

Text #19

Max Range: 0.0055. Min Range: -0.0055

Max Act: 0.0034. Min Act: -0.0002

Data Index: 1293774 (C4 (Web Text))

Max Activating Token Index: 527