Model: SoLU Model: 12 Layers, 6144 Neurons per Layer

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

Neuron 4626 in Layer 10

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-12l')

Text #0

Max Range: 8.2177. Min Range: -8.2177

Max Act: 8.2177. Min Act: -0.0000

Data Index: 656297 (C4 (Web Text))

Max Activating Token Index: 363

Text #1

Max Range: 8.2177. Min Range: -8.2177

Max Act: 8.0863. Min Act: -0.0000

Data Index: 729782 (C4 (Web Text))

Max Activating Token Index: 765

Text #2

Max Range: 8.2177. Min Range: -8.2177

Max Act: 7.0606. Min Act: -0.0000

Data Index: 726084 (C4 (Web Text))

Max Activating Token Index: 285

Text #3

Max Range: 8.2177. Min Range: -8.2177

Max Act: 4.7115. Min Act: -0.0000

Data Index: 118221 (C4 (Web Text))

Max Activating Token Index: 952

Text #4

Max Range: 8.2177. Min Range: -8.2177

Max Act: 4.5621. Min Act: -0.0000

Data Index: 453579 (C4 (Web Text))

Max Activating Token Index: 902

Text #5

Max Range: 8.2177. Min Range: -8.2177

Max Act: 4.3327. Min Act: -0.0000

Data Index: 1095517 (C4 (Web Text))

Max Activating Token Index: 735

Text #6

Max Range: 8.2177. Min Range: -8.2177

Max Act: 4.3206. Min Act: -0.0000

Data Index: 64125 (C4 (Web Text))

Max Activating Token Index: 603

Text #7

Max Range: 8.2177. Min Range: -8.2177

Max Act: 4.7725. Min Act: -0.0000

Data Index: 1437525 (Python Code)

Max Activating Token Index: 198

Text #8

Max Range: 8.2177. Min Range: -8.2177

Max Act: 4.0169. Min Act: -0.0000

Data Index: 1101761 (C4 (Web Text))

Max Activating Token Index: 367

Text #9

Max Range: 8.2177. Min Range: -8.2177

Max Act: 3.3348. Min Act: -0.0000

Data Index: 1600241 (Python Code)

Max Activating Token Index: 413

Text #10

Max Range: 8.2177. Min Range: -8.2177

Max Act: 3.4433. Min Act: -0.0000

Data Index: 624729 (C4 (Web Text))

Max Activating Token Index: 448

Text #11

Max Range: 8.2177. Min Range: -8.2177

Max Act: 3.8995. Min Act: -0.0000

Data Index: 898238 (C4 (Web Text))

Max Activating Token Index: 276

Text #12

Max Range: 8.2177. Min Range: -8.2177

Max Act: 3.2686. Min Act: -0.0000

Data Index: 1017988 (C4 (Web Text))

Max Activating Token Index: 788

Text #13

Max Range: 8.2177. Min Range: -8.2177

Max Act: 3.2757. Min Act: -0.0000

Data Index: 183273 (C4 (Web Text))

Max Activating Token Index: 276

Text #14

Max Range: 8.2177. Min Range: -8.2177

Max Act: 3.4331. Min Act: -0.0000

Data Index: 1329201 (C4 (Web Text))

Max Activating Token Index: 395

Text #15

Max Range: 8.2177. Min Range: -8.2177

Max Act: 4.0810. Min Act: -0.0000

Data Index: 1410781 (Python Code)

Max Activating Token Index: 380

Text #16

Max Range: 8.2177. Min Range: -8.2177

Max Act: 2.9476. Min Act: -0.0000

Data Index: 786537 (C4 (Web Text))

Max Activating Token Index: 334

Text #17

Max Range: 8.2177. Min Range: -8.2177

Max Act: 3.2901. Min Act: -0.0000

Data Index: 1280826 (C4 (Web Text))

Max Activating Token Index: 915

Text #18

Max Range: 8.2177. Min Range: -8.2177

Max Act: 3.4703. Min Act: -0.0000

Data Index: 80480 (C4 (Web Text))

Max Activating Token Index: 917

Text #19

Max Range: 8.2177. Min Range: -8.2177

Max Act: 3.0896. Min Act: -0.0000

Data Index: 1061470 (C4 (Web Text))

Max Activating Token Index: 60