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

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

Neuron 480 in Layer 5

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 1.2853. Min Range: -1.2853

Max Act: 1.2853. Min Act: -0.0001

Data Index: 552269 (C4 (Web Text))

Max Activating Token Index: 447

Text #1

Max Range: 1.2853. Min Range: -1.2853

Max Act: 1.0051. Min Act: -0.0001

Data Index: 1329691 (C4 (Web Text))

Max Activating Token Index: 762

Text #2

Max Range: 1.2853. Min Range: -1.2853

Max Act: 1.1300. Min Act: -0.0000

Data Index: 568292 (C4 (Web Text))

Max Activating Token Index: 986

Text #3

Max Range: 1.2853. Min Range: -1.2853

Max Act: 0.9802. Min Act: -0.0000

Data Index: 313266 (C4 (Web Text))

Max Activating Token Index: 201

Text #4

Max Range: 1.2853. Min Range: -1.2853

Max Act: 0.9391. Min Act: -0.0001

Data Index: 548844 (C4 (Web Text))

Max Activating Token Index: 851

Text #5

Max Range: 1.2853. Min Range: -1.2853

Max Act: 0.9193. Min Act: -0.0001

Data Index: 220102 (C4 (Web Text))

Max Activating Token Index: 169

Text #6

Max Range: 1.2853. Min Range: -1.2853

Max Act: 0.9420. Min Act: -0.0000

Data Index: 329370 (C4 (Web Text))

Max Activating Token Index: 224

Text #7

Max Range: 1.2853. Min Range: -1.2853

Max Act: 0.9954. Min Act: -0.0001

Data Index: 935428 (C4 (Web Text))

Max Activating Token Index: 306

Text #8

Max Range: 1.2853. Min Range: -1.2853

Max Act: 0.8717. Min Act: -0.0000

Data Index: 386443 (C4 (Web Text))

Max Activating Token Index: 365

Text #9

Max Range: 1.2853. Min Range: -1.2853

Max Act: 0.8747. Min Act: -0.0000

Data Index: 260425 (C4 (Web Text))

Max Activating Token Index: 670

Text #10

Max Range: 1.2853. Min Range: -1.2853

Max Act: 0.8469. Min Act: -0.0001

Data Index: 1172461 (C4 (Web Text))

Max Activating Token Index: 258

Text #11

Max Range: 1.2853. Min Range: -1.2853

Max Act: 0.9289. Min Act: -0.0000

Data Index: 1255543 (C4 (Web Text))

Max Activating Token Index: 804

Text #12

Max Range: 1.2853. Min Range: -1.2853

Max Act: 0.9230. Min Act: -0.0000

Data Index: 569925 (C4 (Web Text))

Max Activating Token Index: 971

Text #13

Max Range: 1.2853. Min Range: -1.2853

Max Act: 0.8875. Min Act: -0.0000

Data Index: 1501337 (Python Code)

Max Activating Token Index: 1012

Text #14

Max Range: 1.2853. Min Range: -1.2853

Max Act: 0.8017. Min Act: -0.0001

Data Index: 951867 (C4 (Web Text))

Max Activating Token Index: 750

Text #15

Max Range: 1.2853. Min Range: -1.2853

Max Act: 0.8915. Min Act: -0.0000

Data Index: 1327695 (C4 (Web Text))

Max Activating Token Index: 333

Text #16

Max Range: 1.2853. Min Range: -1.2853

Max Act: 0.8928. Min Act: -0.0001

Data Index: 987104 (C4 (Web Text))

Max Activating Token Index: 703

Text #17

Max Range: 1.2853. Min Range: -1.2853

Max Act: 0.8777. Min Act: -0.0001

Data Index: 1388728 (Python Code)

Max Activating Token Index: 1018

Text #18

Max Range: 1.2853. Min Range: -1.2853

Max Act: 0.8471. Min Act: -0.0000

Data Index: 291125 (C4 (Web Text))

Max Activating Token Index: 319

Text #19

Max Range: 1.2853. Min Range: -1.2853

Max Act: 0.8255. Min Act: -0.0001

Data Index: 277833 (C4 (Web Text))

Max Activating Token Index: 549