Model: SoLU Model: 4 Layers, 2048 Neurons per Layer

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

Neuron 1590 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-4l')

Text #0

Max Range: 0.0673. Min Range: -0.0673

Max Act: 0.0673. Min Act: -0.0002

Data Index: 39448 (C4 (Web Text))

Max Activating Token Index: 65

Text #1

Max Range: 0.0673. Min Range: -0.0673

Max Act: 0.0573. Min Act: -0.0002

Data Index: 344940 (C4 (Web Text))

Max Activating Token Index: 961

Text #2

Max Range: 0.0673. Min Range: -0.0673

Max Act: 0.0559. Min Act: -0.0002

Data Index: 1030806 (C4 (Web Text))

Max Activating Token Index: 406

Text #3

Max Range: 0.0673. Min Range: -0.0673

Max Act: 0.0547. Min Act: -0.0002

Data Index: 404645 (C4 (Web Text))

Max Activating Token Index: 212

Text #4

Max Range: 0.0673. Min Range: -0.0673

Max Act: 0.0564. Min Act: -0.0002

Data Index: 329731 (C4 (Web Text))

Max Activating Token Index: 105

Text #5

Max Range: 0.0673. Min Range: -0.0673

Max Act: 0.0545. Min Act: -0.0002

Data Index: 439971 (C4 (Web Text))

Max Activating Token Index: 241

Text #6

Max Range: 0.0673. Min Range: -0.0673

Max Act: 0.0528. Min Act: -0.0002

Data Index: 624198 (C4 (Web Text))

Max Activating Token Index: 83

Text #7

Max Range: 0.0673. Min Range: -0.0673

Max Act: 0.0529. Min Act: -0.0002

Data Index: 684062 (C4 (Web Text))

Max Activating Token Index: 902

Text #8

Max Range: 0.0673. Min Range: -0.0673

Max Act: 0.0523. Min Act: -0.0002

Data Index: 163022 (C4 (Web Text))

Max Activating Token Index: 633

Text #9

Max Range: 0.0673. Min Range: -0.0673

Max Act: 0.0505. Min Act: -0.0001

Data Index: 953541 (C4 (Web Text))

Max Activating Token Index: 50

Text #10

Max Range: 0.0673. Min Range: -0.0673

Max Act: 0.0500. Min Act: -0.0002

Data Index: 594490 (C4 (Web Text))

Max Activating Token Index: 543

Text #11

Max Range: 0.0673. Min Range: -0.0673

Max Act: 0.0499. Min Act: -0.0002

Data Index: 93288 (C4 (Web Text))

Max Activating Token Index: 803

Text #12

Max Range: 0.0673. Min Range: -0.0673

Max Act: 0.0486. Min Act: -0.0002

Data Index: 311509 (C4 (Web Text))

Max Activating Token Index: 83

Text #13

Max Range: 0.0673. Min Range: -0.0673

Max Act: 0.0488. Min Act: -0.0002

Data Index: 1406484 (Python Code)

Max Activating Token Index: 612

Text #14

Max Range: 0.0673. Min Range: -0.0673

Max Act: 0.0490. Min Act: -0.0002

Data Index: 625467 (C4 (Web Text))

Max Activating Token Index: 27

Text #15

Max Range: 0.0673. Min Range: -0.0673

Max Act: 0.0491. Min Act: -0.0002

Data Index: 576074 (C4 (Web Text))

Max Activating Token Index: 1001

Text #16

Max Range: 0.0673. Min Range: -0.0673

Max Act: 0.0492. Min Act: -0.0002

Data Index: 1579899 (Python Code)

Max Activating Token Index: 1016

Text #17

Max Range: 0.0673. Min Range: -0.0673

Max Act: 0.0485. Min Act: -0.0002

Data Index: 880292 (C4 (Web Text))

Max Activating Token Index: 563

Text #18

Max Range: 0.0673. Min Range: -0.0673

Max Act: 0.0472. Min Act: -0.0002

Data Index: 1141644 (C4 (Web Text))

Max Activating Token Index: 570

Text #19

Max Range: 0.0673. Min Range: -0.0673

Max Act: 0.0481. Min Act: -0.0002

Data Index: 896340 (C4 (Web Text))

Max Activating Token Index: 161