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

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

Neuron 330 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.0732. Min Range: -0.0732

Max Act: 0.0732. Min Act: -0.0006

Data Index: 982600 (C4 (Web Text))

Max Activating Token Index: 482

Text #1

Max Range: 0.0732. Min Range: -0.0732

Max Act: 0.0685. Min Act: -0.0006

Data Index: 466264 (C4 (Web Text))

Max Activating Token Index: 715

Text #2

Max Range: 0.0732. Min Range: -0.0732

Max Act: 0.0682. Min Act: -0.0006

Data Index: 1164805 (C4 (Web Text))

Max Activating Token Index: 536

Text #3

Max Range: 0.0732. Min Range: -0.0732

Max Act: 0.0678. Min Act: -0.0006

Data Index: 874675 (C4 (Web Text))

Max Activating Token Index: 616

Text #4

Max Range: 0.0732. Min Range: -0.0732

Max Act: 0.0646. Min Act: -0.0006

Data Index: 1311331 (C4 (Web Text))

Max Activating Token Index: 278

Text #5

Max Range: 0.0732. Min Range: -0.0732

Max Act: 0.0644. Min Act: -0.0006

Data Index: 1315681 (C4 (Web Text))

Max Activating Token Index: 601

Text #6

Max Range: 0.0732. Min Range: -0.0732

Max Act: 0.0628. Min Act: -0.0006

Data Index: 735728 (C4 (Web Text))

Max Activating Token Index: 492

Text #7

Max Range: 0.0732. Min Range: -0.0732

Max Act: 0.0641. Min Act: -0.0006

Data Index: 279488 (C4 (Web Text))

Max Activating Token Index: 208

Text #8

Max Range: 0.0732. Min Range: -0.0732

Max Act: 0.0638. Min Act: -0.0006

Data Index: 932732 (C4 (Web Text))

Max Activating Token Index: 767

Text #9

Max Range: 0.0732. Min Range: -0.0732

Max Act: 0.0636. Min Act: -0.0006

Data Index: 569223 (C4 (Web Text))

Max Activating Token Index: 160

Text #10

Max Range: 0.0732. Min Range: -0.0732

Max Act: 0.0633. Min Act: -0.0006

Data Index: 955230 (C4 (Web Text))

Max Activating Token Index: 451

Text #11

Max Range: 0.0732. Min Range: -0.0732

Max Act: 0.0648. Min Act: -0.0006

Data Index: 981707 (C4 (Web Text))

Max Activating Token Index: 763

Text #12

Max Range: 0.0732. Min Range: -0.0732

Max Act: 0.0632. Min Act: -0.0006

Data Index: 152637 (C4 (Web Text))

Max Activating Token Index: 241

Text #13

Max Range: 0.0732. Min Range: -0.0732

Max Act: 0.0644. Min Act: -0.0006

Data Index: 472560 (C4 (Web Text))

Max Activating Token Index: 294

Text #14

Max Range: 0.0732. Min Range: -0.0732

Max Act: 0.0616. Min Act: -0.0006

Data Index: 1494811 (Python Code)

Max Activating Token Index: 49

Text #15

Max Range: 0.0732. Min Range: -0.0732

Max Act: 0.0626. Min Act: -0.0006

Data Index: 893505 (C4 (Web Text))

Max Activating Token Index: 307

Text #16

Max Range: 0.0732. Min Range: -0.0732

Max Act: 0.0612. Min Act: -0.0006

Data Index: 1419759 (Python Code)

Max Activating Token Index: 788

Text #17

Max Range: 0.0732. Min Range: -0.0732

Max Act: 0.0622. Min Act: -0.0006

Data Index: 1294221 (C4 (Web Text))

Max Activating Token Index: 535

Text #18

Max Range: 0.0732. Min Range: -0.0732

Max Act: 0.0606. Min Act: -0.0006

Data Index: 123090 (C4 (Web Text))

Max Activating Token Index: 440

Text #19

Max Range: 0.0732. Min Range: -0.0732

Max Act: 0.0613. Min Act: -0.0006

Data Index: 1209178 (C4 (Web Text))

Max Activating Token Index: 506