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

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

Neuron 1913 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 1.3568. Min Range: -1.3568

Max Act: 1.3568. Min Act: -0.0001

Data Index: 395670 (C4 (Web Text))

Max Activating Token Index: 907

Text #1

Max Range: 1.3568. Min Range: -1.3568

Max Act: 1.1363. Min Act: -0.0001

Data Index: 1653171 (Python Code)

Max Activating Token Index: 437

Text #2

Max Range: 1.3568. Min Range: -1.3568

Max Act: 1.0150. Min Act: -0.0001

Data Index: 987696 (C4 (Web Text))

Max Activating Token Index: 365

Text #3

Max Range: 1.3568. Min Range: -1.3568

Max Act: 1.0110. Min Act: -0.0001

Data Index: 1288994 (C4 (Web Text))

Max Activating Token Index: 837

Text #4

Max Range: 1.3568. Min Range: -1.3568

Max Act: 1.0017. Min Act: -0.0001

Data Index: 605428 (C4 (Web Text))

Max Activating Token Index: 270

Text #5

Max Range: 1.3568. Min Range: -1.3568

Max Act: 0.9857. Min Act: -0.0001

Data Index: 55439 (C4 (Web Text))

Max Activating Token Index: 9

Text #6

Max Range: 1.3568. Min Range: -1.3568

Max Act: 0.9291. Min Act: -0.0001

Data Index: 798918 (C4 (Web Text))

Max Activating Token Index: 414

Text #7

Max Range: 1.3568. Min Range: -1.3568

Max Act: 1.0157. Min Act: -0.0001

Data Index: 262411 (C4 (Web Text))

Max Activating Token Index: 407

Text #8

Max Range: 1.3568. Min Range: -1.3568

Max Act: 0.9678. Min Act: -0.0001

Data Index: 1019866 (C4 (Web Text))

Max Activating Token Index: 729

Text #9

Max Range: 1.3568. Min Range: -1.3568

Max Act: 1.0009. Min Act: -0.0001

Data Index: 1360252 (Python Code)

Max Activating Token Index: 987

Text #10

Max Range: 1.3568. Min Range: -1.3568

Max Act: 0.9592. Min Act: -0.0001

Data Index: 1546379 (Python Code)

Max Activating Token Index: 14

Text #11

Max Range: 1.3568. Min Range: -1.3568

Max Act: 0.9760. Min Act: -0.0001

Data Index: 1644620 (Python Code)

Max Activating Token Index: 985

Text #12

Max Range: 1.3568. Min Range: -1.3568

Max Act: 0.8714. Min Act: -0.0001

Data Index: 1049282 (C4 (Web Text))

Max Activating Token Index: 795

Text #13

Max Range: 1.3568. Min Range: -1.3568

Max Act: 0.9925. Min Act: -0.0001

Data Index: 1399436 (Python Code)

Max Activating Token Index: 606

Text #14

Max Range: 1.3568. Min Range: -1.3568

Max Act: 0.9433. Min Act: -0.0001

Data Index: 1054946 (C4 (Web Text))

Max Activating Token Index: 562

Text #15

Max Range: 1.3568. Min Range: -1.3568

Max Act: 0.9706. Min Act: -0.0001

Data Index: 1561390 (Python Code)

Max Activating Token Index: 698

Text #16

Max Range: 1.3568. Min Range: -1.3568

Max Act: 0.8922. Min Act: -0.0001

Data Index: 497314 (C4 (Web Text))

Max Activating Token Index: 365

Text #17

Max Range: 1.3568. Min Range: -1.3568

Max Act: 1.0218. Min Act: -0.0001

Data Index: 597473 (C4 (Web Text))

Max Activating Token Index: 304

Text #18

Max Range: 1.3568. Min Range: -1.3568

Max Act: 0.9705. Min Act: -0.0001

Data Index: 901308 (C4 (Web Text))

Max Activating Token Index: 632

Text #19

Max Range: 1.3568. Min Range: -1.3568

Max Act: 0.9971. Min Act: -0.0001

Data Index: 65795 (C4 (Web Text))

Max Activating Token Index: 724