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

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

Neuron 2721 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: 1.2892. Min Range: -1.2892

Max Act: 1.2892. Min Act: -0.0000

Data Index: 58858 (C4 (Web Text))

Max Activating Token Index: 183

Text #1

Max Range: 1.2892. Min Range: -1.2892

Max Act: 1.2119. Min Act: -0.0000

Data Index: 723227 (C4 (Web Text))

Max Activating Token Index: 158

Text #2

Max Range: 1.2892. Min Range: -1.2892

Max Act: 1.2211. Min Act: -0.0000

Data Index: 976708 (C4 (Web Text))

Max Activating Token Index: 337

Text #3

Max Range: 1.2892. Min Range: -1.2892

Max Act: 1.1491. Min Act: -0.0000

Data Index: 1461317 (Python Code)

Max Activating Token Index: 64

Text #4

Max Range: 1.2892. Min Range: -1.2892

Max Act: 0.9755. Min Act: -0.0000

Data Index: 1321680 (C4 (Web Text))

Max Activating Token Index: 293

Text #5

Max Range: 1.2892. Min Range: -1.2892

Max Act: 1.1196. Min Act: -0.0000

Data Index: 1406030 (Python Code)

Max Activating Token Index: 963

Text #6

Max Range: 1.2892. Min Range: -1.2892

Max Act: 1.0352. Min Act: -0.0000

Data Index: 1066785 (C4 (Web Text))

Max Activating Token Index: 308

Text #7

Max Range: 1.2892. Min Range: -1.2892

Max Act: 0.9589. Min Act: -0.0000

Data Index: 659913 (C4 (Web Text))

Max Activating Token Index: 810

Text #8

Max Range: 1.2892. Min Range: -1.2892

Max Act: 0.9396. Min Act: -0.0000

Data Index: 793504 (C4 (Web Text))

Max Activating Token Index: 473

Text #9

Max Range: 1.2892. Min Range: -1.2892

Max Act: 0.9554. Min Act: -0.0000

Data Index: 1555140 (Python Code)

Max Activating Token Index: 847

Text #10

Max Range: 1.2892. Min Range: -1.2892

Max Act: 0.9357. Min Act: 0.0000

Data Index: 177159 (C4 (Web Text))

Max Activating Token Index: 297

Text #11

Max Range: 1.2892. Min Range: -1.2892

Max Act: 0.8873. Min Act: -0.0000

Data Index: 1563285 (Python Code)

Max Activating Token Index: 689

Text #12

Max Range: 1.2892. Min Range: -1.2892

Max Act: 0.9573. Min Act: -0.0000

Data Index: 1588296 (Python Code)

Max Activating Token Index: 39

Text #13

Max Range: 1.2892. Min Range: -1.2892

Max Act: 0.9796. Min Act: -0.0000

Data Index: 1393421 (Python Code)

Max Activating Token Index: 153

Text #14

Max Range: 1.2892. Min Range: -1.2892

Max Act: 0.9020. Min Act: -0.0000

Data Index: 1083058 (C4 (Web Text))

Max Activating Token Index: 277

Text #15

Max Range: 1.2892. Min Range: -1.2892

Max Act: 0.9350. Min Act: 0.0000

Data Index: 1002669 (C4 (Web Text))

Max Activating Token Index: 979

Text #16

Max Range: 1.2892. Min Range: -1.2892

Max Act: 0.9274. Min Act: -0.0000

Data Index: 21696 (C4 (Web Text))

Max Activating Token Index: 453

Text #17

Max Range: 1.2892. Min Range: -1.2892

Max Act: 0.8757. Min Act: -0.0000

Data Index: 1082663 (C4 (Web Text))

Max Activating Token Index: 176

Text #18

Max Range: 1.2892. Min Range: -1.2892

Max Act: 0.8637. Min Act: -0.0000

Data Index: 1623168 (Python Code)

Max Activating Token Index: 475

Text #19

Max Range: 1.2892. Min Range: -1.2892

Max Act: 0.9166. Min Act: -0.0000

Data Index: 1335930 (C4 (Web Text))

Max Activating Token Index: 397