Model: SoLU Model: 10 Layers, 5120 Neurons per Layer

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

Neuron 4851 in Layer 7

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-10l')

Text #0

Max Range: 5.3676. Min Range: -5.3676

Max Act: 5.3676. Min Act: -0.0000

Data Index: 409080 (C4 (Web Text))

Max Activating Token Index: 985

Text #1

Max Range: 5.3676. Min Range: -5.3676

Max Act: 5.3650. Min Act: -0.0000

Data Index: 451182 (C4 (Web Text))

Max Activating Token Index: 398

Text #2

Max Range: 5.3676. Min Range: -5.3676

Max Act: 4.9657. Min Act: -0.0000

Data Index: 1209534 (C4 (Web Text))

Max Activating Token Index: 456

Text #3

Max Range: 5.3676. Min Range: -5.3676

Max Act: 5.1550. Min Act: -0.0000

Data Index: 455083 (C4 (Web Text))

Max Activating Token Index: 794

Text #4

Max Range: 5.3676. Min Range: -5.3676

Max Act: 4.6045. Min Act: -0.0000

Data Index: 774632 (C4 (Web Text))

Max Activating Token Index: 598

Text #5

Max Range: 5.3676. Min Range: -5.3676

Max Act: 4.6375. Min Act: -0.0000

Data Index: 540865 (C4 (Web Text))

Max Activating Token Index: 972

Text #6

Max Range: 5.3676. Min Range: -5.3676

Max Act: 4.7992. Min Act: -0.0000

Data Index: 510022 (C4 (Web Text))

Max Activating Token Index: 43

Text #7

Max Range: 5.3676. Min Range: -5.3676

Max Act: 4.3429. Min Act: -0.0000

Data Index: 1530635 (Python Code)

Max Activating Token Index: 42

Text #8

Max Range: 5.3676. Min Range: -5.3676

Max Act: 4.2184. Min Act: -0.0000

Data Index: 589011 (C4 (Web Text))

Max Activating Token Index: 104

Text #9

Max Range: 5.3676. Min Range: -5.3676

Max Act: 4.3073. Min Act: -0.0000

Data Index: 1138787 (C4 (Web Text))

Max Activating Token Index: 409

Text #10

Max Range: 5.3676. Min Range: -5.3676

Max Act: 4.0096. Min Act: -0.0000

Data Index: 677446 (C4 (Web Text))

Max Activating Token Index: 90

Text #11

Max Range: 5.3676. Min Range: -5.3676

Max Act: 4.2146. Min Act: -0.0000

Data Index: 1502180 (Python Code)

Max Activating Token Index: 346

Text #12

Max Range: 5.3676. Min Range: -5.3676

Max Act: 4.1976. Min Act: -0.0000

Data Index: 419793 (C4 (Web Text))

Max Activating Token Index: 342

Text #13

Max Range: 5.3676. Min Range: -5.3676

Max Act: 3.6632. Min Act: -0.0000

Data Index: 1286625 (C4 (Web Text))

Max Activating Token Index: 606

Text #14

Max Range: 5.3676. Min Range: -5.3676

Max Act: 3.7559. Min Act: -0.0000

Data Index: 1622909 (Python Code)

Max Activating Token Index: 25

Text #15

Max Range: 5.3676. Min Range: -5.3676

Max Act: 3.9164. Min Act: -0.0000

Data Index: 397605 (C4 (Web Text))

Max Activating Token Index: 273

Text #16

Max Range: 5.3676. Min Range: -5.3676

Max Act: 4.0365. Min Act: -0.0000

Data Index: 782939 (C4 (Web Text))

Max Activating Token Index: 845

Text #17

Max Range: 5.3676. Min Range: -5.3676

Max Act: 3.9914. Min Act: -0.0000

Data Index: 608941 (C4 (Web Text))

Max Activating Token Index: 135

Text #18

Max Range: 5.3676. Min Range: -5.3676

Max Act: 3.8140. Min Act: -0.0000

Data Index: 1638861 (Python Code)

Max Activating Token Index: 818

Text #19

Max Range: 5.3676. Min Range: -5.3676

Max Act: 3.8050. Min Act: -0.0000

Data Index: 1331421 (C4 (Web Text))

Max Activating Token Index: 801