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

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

Neuron 1428 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.0888. Min Range: -0.0888

Max Act: 0.0888. Min Act: -0.0001

Data Index: 883764 (C4 (Web Text))

Max Activating Token Index: 86

Text #1

Max Range: 0.0888. Min Range: -0.0888

Max Act: 0.0785. Min Act: -0.0001

Data Index: 589440 (C4 (Web Text))

Max Activating Token Index: 597

Text #2

Max Range: 0.0888. Min Range: -0.0888

Max Act: 0.0737. Min Act: -0.0001

Data Index: 1139384 (C4 (Web Text))

Max Activating Token Index: 433

Text #3

Max Range: 0.0888. Min Range: -0.0888

Max Act: 0.0786. Min Act: -0.0001

Data Index: 360802 (C4 (Web Text))

Max Activating Token Index: 142

Text #4

Max Range: 0.0888. Min Range: -0.0888

Max Act: 0.0767. Min Act: -0.0001

Data Index: 1319022 (C4 (Web Text))

Max Activating Token Index: 448

Text #5

Max Range: 0.0888. Min Range: -0.0888

Max Act: 0.0758. Min Act: -0.0001

Data Index: 1190548 (C4 (Web Text))

Max Activating Token Index: 298

Text #6

Max Range: 0.0888. Min Range: -0.0888

Max Act: 0.0703. Min Act: -0.0001

Data Index: 377193 (C4 (Web Text))

Max Activating Token Index: 21

Text #7

Max Range: 0.0888. Min Range: -0.0888

Max Act: 0.0768. Min Act: -0.0001

Data Index: 1145247 (C4 (Web Text))

Max Activating Token Index: 407

Text #8

Max Range: 0.0888. Min Range: -0.0888

Max Act: 0.0708. Min Act: -0.0002

Data Index: 135484 (C4 (Web Text))

Max Activating Token Index: 6

Text #9

Max Range: 0.0888. Min Range: -0.0888

Max Act: 0.0708. Min Act: -0.0001

Data Index: 1183063 (C4 (Web Text))

Max Activating Token Index: 6

Text #10

Max Range: 0.0888. Min Range: -0.0888

Max Act: 0.0728. Min Act: -0.0001

Data Index: 1586612 (Python Code)

Max Activating Token Index: 450

Text #11

Max Range: 0.0888. Min Range: -0.0888

Max Act: 0.0769. Min Act: -0.0001

Data Index: 1487183 (Python Code)

Max Activating Token Index: 1021

Text #12

Max Range: 0.0888. Min Range: -0.0888

Max Act: 0.0645. Min Act: -0.0001

Data Index: 1412100 (Python Code)

Max Activating Token Index: 663

Text #13

Max Range: 0.0888. Min Range: -0.0888

Max Act: 0.0612. Min Act: -0.0001

Data Index: 305576 (C4 (Web Text))

Max Activating Token Index: 996

Text #14

Max Range: 0.0888. Min Range: -0.0888

Max Act: 0.0676. Min Act: -0.0001

Data Index: 1404280 (Python Code)

Max Activating Token Index: 999

Text #15

Max Range: 0.0888. Min Range: -0.0888

Max Act: 0.0645. Min Act: -0.0001

Data Index: 706226 (C4 (Web Text))

Max Activating Token Index: 163

Text #16

Max Range: 0.0888. Min Range: -0.0888

Max Act: 0.0658. Min Act: -0.0001

Data Index: 1459924 (Python Code)

Max Activating Token Index: 279

Text #17

Max Range: 0.0888. Min Range: -0.0888

Max Act: 0.0585. Min Act: -0.0002

Data Index: 769872 (C4 (Web Text))

Max Activating Token Index: 905

Text #18

Max Range: 0.0888. Min Range: -0.0888

Max Act: 0.0654. Min Act: -0.0001

Data Index: 519379 (C4 (Web Text))

Max Activating Token Index: 14

Text #19

Max Range: 0.0888. Min Range: -0.0888

Max Act: 0.0649. Min Act: -0.0001

Data Index: 245930 (C4 (Web Text))

Max Activating Token Index: 73