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

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

Neuron 3737 in Layer 5

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.4907. Min Range: -0.4907

Max Act: 0.4907. Min Act: -0.0001

Data Index: 1422281 (Python Code)

Max Activating Token Index: 72

Text #1

Max Range: 0.4907. Min Range: -0.4907

Max Act: 0.3836. Min Act: -0.0001

Data Index: 1195564 (C4 (Web Text))

Max Activating Token Index: 205

Text #2

Max Range: 0.4907. Min Range: -0.4907

Max Act: 0.3879. Min Act: -0.0001

Data Index: 280529 (C4 (Web Text))

Max Activating Token Index: 774

Text #3

Max Range: 0.4907. Min Range: -0.4907

Max Act: 0.3307. Min Act: -0.0001

Data Index: 222975 (C4 (Web Text))

Max Activating Token Index: 805

Text #4

Max Range: 0.4907. Min Range: -0.4907

Max Act: 0.3400. Min Act: -0.0001

Data Index: 113051 (C4 (Web Text))

Max Activating Token Index: 95

Text #5

Max Range: 0.4907. Min Range: -0.4907

Max Act: 0.3154. Min Act: -0.0001

Data Index: 1026349 (C4 (Web Text))

Max Activating Token Index: 719

Text #6

Max Range: 0.4907. Min Range: -0.4907

Max Act: 0.3361. Min Act: -0.0001

Data Index: 1009806 (C4 (Web Text))

Max Activating Token Index: 137

Text #7

Max Range: 0.4907. Min Range: -0.4907

Max Act: 0.3075. Min Act: -0.0001

Data Index: 319600 (C4 (Web Text))

Max Activating Token Index: 506

Text #8

Max Range: 0.4907. Min Range: -0.4907

Max Act: 0.3398. Min Act: -0.0001

Data Index: 1077536 (C4 (Web Text))

Max Activating Token Index: 87

Text #9

Max Range: 0.4907. Min Range: -0.4907

Max Act: 0.3042. Min Act: -0.0001

Data Index: 419581 (C4 (Web Text))

Max Activating Token Index: 374

Text #10

Max Range: 0.4907. Min Range: -0.4907

Max Act: 0.3048. Min Act: -0.0001

Data Index: 703692 (C4 (Web Text))

Max Activating Token Index: 171

Text #11

Max Range: 0.4907. Min Range: -0.4907

Max Act: 0.2819. Min Act: -0.0001

Data Index: 2402 (C4 (Web Text))

Max Activating Token Index: 600

Text #12

Max Range: 0.4907. Min Range: -0.4907

Max Act: 0.2911. Min Act: -0.0001

Data Index: 289400 (C4 (Web Text))

Max Activating Token Index: 264

Text #13

Max Range: 0.4907. Min Range: -0.4907

Max Act: 0.2968. Min Act: -0.0001

Data Index: 279052 (C4 (Web Text))

Max Activating Token Index: 162

Text #14

Max Range: 0.4907. Min Range: -0.4907

Max Act: 0.2877. Min Act: -0.0001

Data Index: 1223293 (C4 (Web Text))

Max Activating Token Index: 563

Text #15

Max Range: 0.4907. Min Range: -0.4907

Max Act: 0.3010. Min Act: -0.0001

Data Index: 907613 (C4 (Web Text))

Max Activating Token Index: 347

Text #16

Max Range: 0.4907. Min Range: -0.4907

Max Act: 0.2882. Min Act: -0.0001

Data Index: 394001 (C4 (Web Text))

Max Activating Token Index: 705

Text #17

Max Range: 0.4907. Min Range: -0.4907

Max Act: 0.2935. Min Act: -0.0001

Data Index: 810367 (C4 (Web Text))

Max Activating Token Index: 1021

Text #18

Max Range: 0.4907. Min Range: -0.4907

Max Act: 0.2943. Min Act: -0.0001

Data Index: 1416731 (Python Code)

Max Activating Token Index: 735

Text #19

Max Range: 0.4907. Min Range: -0.4907

Max Act: 0.2770. Min Act: -0.0001

Data Index: 1004637 (C4 (Web Text))

Max Activating Token Index: 405