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

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

Neuron 683 in Layer 7

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.4465. Min Range: -0.4465

Max Act: 0.4465. Min Act: -0.0000

Data Index: 941809 (C4 (Web Text))

Max Activating Token Index: 851

Text #1

Max Range: 0.4465. Min Range: -0.4465

Max Act: 0.3460. Min Act: -0.0000

Data Index: 923490 (C4 (Web Text))

Max Activating Token Index: 903

Text #2

Max Range: 0.4465. Min Range: -0.4465

Max Act: 0.3652. Min Act: -0.0000

Data Index: 399002 (C4 (Web Text))

Max Activating Token Index: 412

Text #3

Max Range: 0.4465. Min Range: -0.4465

Max Act: 0.3478. Min Act: -0.0000

Data Index: 446864 (C4 (Web Text))

Max Activating Token Index: 612

Text #4

Max Range: 0.4465. Min Range: -0.4465

Max Act: 0.3562. Min Act: -0.0000

Data Index: 702239 (C4 (Web Text))

Max Activating Token Index: 508

Text #5

Max Range: 0.4465. Min Range: -0.4465

Max Act: 0.3121. Min Act: -0.0000

Data Index: 1340027 (C4 (Web Text))

Max Activating Token Index: 392

Text #6

Max Range: 0.4465. Min Range: -0.4465

Max Act: 0.3280. Min Act: -0.0000

Data Index: 1262661 (C4 (Web Text))

Max Activating Token Index: 583

Text #7

Max Range: 0.4465. Min Range: -0.4465

Max Act: 0.3252. Min Act: -0.0000

Data Index: 49022 (C4 (Web Text))

Max Activating Token Index: 258

Text #8

Max Range: 0.4465. Min Range: -0.4465

Max Act: 0.3001. Min Act: -0.0000

Data Index: 363987 (C4 (Web Text))

Max Activating Token Index: 649

Text #9

Max Range: 0.4465. Min Range: -0.4465

Max Act: 0.2983. Min Act: -0.0000

Data Index: 688204 (C4 (Web Text))

Max Activating Token Index: 600

Text #10

Max Range: 0.4465. Min Range: -0.4465

Max Act: 0.2835. Min Act: -0.0000

Data Index: 1327353 (C4 (Web Text))

Max Activating Token Index: 916

Text #11

Max Range: 0.4465. Min Range: -0.4465

Max Act: 0.2916. Min Act: -0.0000

Data Index: 88962 (C4 (Web Text))

Max Activating Token Index: 350

Text #12

Max Range: 0.4465. Min Range: -0.4465

Max Act: 0.3175. Min Act: -0.0000

Data Index: 1419823 (Python Code)

Max Activating Token Index: 655

Text #13

Max Range: 0.4465. Min Range: -0.4465

Max Act: 0.2544. Min Act: -0.0000

Data Index: 718660 (C4 (Web Text))

Max Activating Token Index: 148

Text #14

Max Range: 0.4465. Min Range: -0.4465

Max Act: 0.2744. Min Act: -0.0000

Data Index: 860270 (C4 (Web Text))

Max Activating Token Index: 770

Text #15

Max Range: 0.4465. Min Range: -0.4465

Max Act: 0.2433. Min Act: -0.0000

Data Index: 566476 (C4 (Web Text))

Max Activating Token Index: 858

Text #16

Max Range: 0.4465. Min Range: -0.4465

Max Act: 0.2524. Min Act: -0.0000

Data Index: 906572 (C4 (Web Text))

Max Activating Token Index: 520

Text #17

Max Range: 0.4465. Min Range: -0.4465

Max Act: 0.2182. Min Act: -0.0000

Data Index: 208451 (C4 (Web Text))

Max Activating Token Index: 898

Text #18

Max Range: 0.4465. Min Range: -0.4465

Max Act: 0.2742. Min Act: -0.0000

Data Index: 963883 (C4 (Web Text))

Max Activating Token Index: 832

Text #19

Max Range: 0.4465. Min Range: -0.4465

Max Act: 0.2204. Min Act: -0.0000

Data Index: 520032 (C4 (Web Text))

Max Activating Token Index: 760