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

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

Neuron 920 in Layer 5

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.5903. Min Range: -0.5903

Max Act: 0.5903. Min Act: -0.0000

Data Index: 14140 (C4 (Web Text))

Max Activating Token Index: 99

Text #1

Max Range: 0.5903. Min Range: -0.5903

Max Act: 0.5578. Min Act: -0.0000

Data Index: 1031017 (C4 (Web Text))

Max Activating Token Index: 757

Text #2

Max Range: 0.5903. Min Range: -0.5903

Max Act: 0.5238. Min Act: -0.0000

Data Index: 735509 (C4 (Web Text))

Max Activating Token Index: 652

Text #3

Max Range: 0.5903. Min Range: -0.5903

Max Act: 0.5386. Min Act: -0.0000

Data Index: 165824 (C4 (Web Text))

Max Activating Token Index: 193

Text #4

Max Range: 0.5903. Min Range: -0.5903

Max Act: 0.4530. Min Act: -0.0000

Data Index: 1528990 (Python Code)

Max Activating Token Index: 63

Text #5

Max Range: 0.5903. Min Range: -0.5903

Max Act: 0.4136. Min Act: -0.0000

Data Index: 812086 (C4 (Web Text))

Max Activating Token Index: 369

Text #6

Max Range: 0.5903. Min Range: -0.5903

Max Act: 0.4480. Min Act: -0.0000

Data Index: 31707 (C4 (Web Text))

Max Activating Token Index: 118

Text #7

Max Range: 0.5903. Min Range: -0.5903

Max Act: 0.4643. Min Act: -0.0000

Data Index: 552786 (C4 (Web Text))

Max Activating Token Index: 384

Text #8

Max Range: 0.5903. Min Range: -0.5903

Max Act: 0.4088. Min Act: -0.0000

Data Index: 1296691 (C4 (Web Text))

Max Activating Token Index: 373

Text #9

Max Range: 0.5903. Min Range: -0.5903

Max Act: 0.3706. Min Act: -0.0000

Data Index: 1025080 (C4 (Web Text))

Max Activating Token Index: 224

Text #10

Max Range: 0.5903. Min Range: -0.5903

Max Act: 0.4224. Min Act: -0.0000

Data Index: 1576492 (Python Code)

Max Activating Token Index: 95

Text #11

Max Range: 0.5903. Min Range: -0.5903

Max Act: 0.4122. Min Act: -0.0000

Data Index: 1316357 (C4 (Web Text))

Max Activating Token Index: 917

Text #12

Max Range: 0.5903. Min Range: -0.5903

Max Act: 0.4032. Min Act: -0.0000

Data Index: 1129299 (C4 (Web Text))

Max Activating Token Index: 110

Text #13

Max Range: 0.5903. Min Range: -0.5903

Max Act: 0.4394. Min Act: -0.0000

Data Index: 1064028 (C4 (Web Text))

Max Activating Token Index: 136

Text #14

Max Range: 0.5903. Min Range: -0.5903

Max Act: 0.3699. Min Act: -0.0000

Data Index: 1441056 (Python Code)

Max Activating Token Index: 404

Text #15

Max Range: 0.5903. Min Range: -0.5903

Max Act: 0.4141. Min Act: -0.0000

Data Index: 35905 (C4 (Web Text))

Max Activating Token Index: 403

Text #16

Max Range: 0.5903. Min Range: -0.5903

Max Act: 0.4313. Min Act: -0.0000

Data Index: 1108348 (C4 (Web Text))

Max Activating Token Index: 815

Text #17

Max Range: 0.5903. Min Range: -0.5903

Max Act: 0.3890. Min Act: -0.0000

Data Index: 232461 (C4 (Web Text))

Max Activating Token Index: 361

Text #18

Max Range: 0.5903. Min Range: -0.5903

Max Act: 0.3968. Min Act: -0.0000

Data Index: 315050 (C4 (Web Text))

Max Activating Token Index: 839

Text #19

Max Range: 0.5903. Min Range: -0.5903

Max Act: 0.4051. Min Act: -0.0000

Data Index: 888941 (C4 (Web Text))

Max Activating Token Index: 719