Model: SoLU Model: 3 Layers, 2048 Neurons per Layer

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

Neuron 113 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-3l')

Text #0

Max Range: 0.0622. Min Range: -0.0622

Max Act: 0.0622. Min Act: -0.0003

Data Index: 183675 (C4 (Web Text))

Max Activating Token Index: 267

Text #1

Max Range: 0.0622. Min Range: -0.0622

Max Act: 0.0613. Min Act: -0.0003

Data Index: 205162 (C4 (Web Text))

Max Activating Token Index: 599

Text #2

Max Range: 0.0622. Min Range: -0.0622

Max Act: 0.0542. Min Act: -0.0003

Data Index: 1121057 (C4 (Web Text))

Max Activating Token Index: 1005

Text #3

Max Range: 0.0622. Min Range: -0.0622

Max Act: 0.0534. Min Act: -0.0003

Data Index: 610032 (C4 (Web Text))

Max Activating Token Index: 886

Text #4

Max Range: 0.0622. Min Range: -0.0622

Max Act: 0.0536. Min Act: -0.0003

Data Index: 1256021 (C4 (Web Text))

Max Activating Token Index: 383

Text #5

Max Range: 0.0622. Min Range: -0.0622

Max Act: 0.0491. Min Act: -0.0003

Data Index: 897631 (C4 (Web Text))

Max Activating Token Index: 565

Text #6

Max Range: 0.0622. Min Range: -0.0622

Max Act: 0.0505. Min Act: -0.0003

Data Index: 1532631 (Python Code)

Max Activating Token Index: 789

Text #7

Max Range: 0.0622. Min Range: -0.0622

Max Act: 0.0520. Min Act: -0.0003

Data Index: 1118161 (C4 (Web Text))

Max Activating Token Index: 57

Text #8

Max Range: 0.0622. Min Range: -0.0622

Max Act: 0.0505. Min Act: -0.0003

Data Index: 598283 (C4 (Web Text))

Max Activating Token Index: 52

Text #9

Max Range: 0.0622. Min Range: -0.0622

Max Act: 0.0498. Min Act: -0.0003

Data Index: 1585250 (Python Code)

Max Activating Token Index: 450

Text #10

Max Range: 0.0622. Min Range: -0.0622

Max Act: 0.0495. Min Act: -0.0003

Data Index: 1634183 (Python Code)

Max Activating Token Index: 699

Text #11

Max Range: 0.0622. Min Range: -0.0622

Max Act: 0.0510. Min Act: -0.0003

Data Index: 1564336 (Python Code)

Max Activating Token Index: 465

Text #12

Max Range: 0.0622. Min Range: -0.0622

Max Act: 0.0487. Min Act: -0.0003

Data Index: 320942 (C4 (Web Text))

Max Activating Token Index: 150

Text #13

Max Range: 0.0622. Min Range: -0.0622

Max Act: 0.0486. Min Act: -0.0003

Data Index: 186133 (C4 (Web Text))

Max Activating Token Index: 41

Text #14

Max Range: 0.0622. Min Range: -0.0622

Max Act: 0.0501. Min Act: -0.0003

Data Index: 1431366 (Python Code)

Max Activating Token Index: 338

Text #15

Max Range: 0.0622. Min Range: -0.0622

Max Act: 0.0482. Min Act: -0.0003

Data Index: 636378 (C4 (Web Text))

Max Activating Token Index: 349

Text #16

Max Range: 0.0622. Min Range: -0.0622

Max Act: 0.0463. Min Act: -0.0003

Data Index: 354440 (C4 (Web Text))

Max Activating Token Index: 335

Text #17

Max Range: 0.0622. Min Range: -0.0622

Max Act: 0.0484. Min Act: -0.0003

Data Index: 94728 (C4 (Web Text))

Max Activating Token Index: 59

Text #18

Max Range: 0.0622. Min Range: -0.0622

Max Act: 0.0475. Min Act: -0.0002

Data Index: 697812 (C4 (Web Text))

Max Activating Token Index: 321

Text #19

Max Range: 0.0622. Min Range: -0.0622

Max Act: 0.0470. Min Act: -0.0003

Data Index: 656921 (C4 (Web Text))

Max Activating Token Index: 606