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

Dataset: The Pile

Neuron 1024 in Layer 6

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 7.6724. Min Range: -7.6724

Max Act: 7.6724. Min Act: -0.0000

Data Index: 557069 (The Pile)

Max Activating Token Index: 41

Text #1

Max Range: 7.6724. Min Range: -7.6724

Max Act: 7.1241. Min Act: -0.0000

Data Index: 613269 (The Pile)

Max Activating Token Index: 750

Text #2

Max Range: 7.6724. Min Range: -7.6724

Max Act: 7.2091. Min Act: -0.0000

Data Index: 485444 (The Pile)

Max Activating Token Index: 690

Text #3

Max Range: 7.6724. Min Range: -7.6724

Max Act: 7.0659. Min Act: -0.0000

Data Index: 1812703 (The Pile)

Max Activating Token Index: 216

Text #4

Max Range: 7.6724. Min Range: -7.6724

Max Act: 7.1605. Min Act: -0.0000

Data Index: 1412376 (The Pile)

Max Activating Token Index: 611

Text #5

Max Range: 7.6724. Min Range: -7.6724

Max Act: 6.9705. Min Act: -0.0000

Data Index: 1866411 (The Pile)

Max Activating Token Index: 509

Text #6

Max Range: 7.6724. Min Range: -7.6724

Max Act: 6.7575. Min Act: -0.0000

Data Index: 98748 (The Pile)

Max Activating Token Index: 597

Text #7

Max Range: 7.6724. Min Range: -7.6724

Max Act: 6.8901. Min Act: -0.0000

Data Index: 1870807 (The Pile)

Max Activating Token Index: 462

Text #8

Max Range: 7.6724. Min Range: -7.6724

Max Act: 6.6772. Min Act: -0.0000

Data Index: 448251 (The Pile)

Max Activating Token Index: 631

Text #9

Max Range: 7.6724. Min Range: -7.6724

Max Act: 6.8189. Min Act: -0.0000

Data Index: 1311911 (The Pile)

Max Activating Token Index: 582

Text #10

Max Range: 7.6724. Min Range: -7.6724

Max Act: 6.9787. Min Act: -0.0000

Data Index: 1191821 (The Pile)

Max Activating Token Index: 280

Text #11

Max Range: 7.6724. Min Range: -7.6724

Max Act: 6.9156. Min Act: -0.0000

Data Index: 1126911 (The Pile)

Max Activating Token Index: 508

Text #12

Max Range: 7.6724. Min Range: -7.6724

Max Act: 7.0683. Min Act: -0.0000

Data Index: 1009797 (The Pile)

Max Activating Token Index: 1005

Text #13

Max Range: 7.6724. Min Range: -7.6724

Max Act: 6.5948. Min Act: -0.0000

Data Index: 40100 (The Pile)

Max Activating Token Index: 143

Text #14

Max Range: 7.6724. Min Range: -7.6724

Max Act: 6.7498. Min Act: -0.0000

Data Index: 999030 (The Pile)

Max Activating Token Index: 526

Text #15

Max Range: 7.6724. Min Range: -7.6724

Max Act: 6.5081. Min Act: -0.0000

Data Index: 969852 (The Pile)

Max Activating Token Index: 566

Text #16

Max Range: 7.6724. Min Range: -7.6724

Max Act: 6.3009. Min Act: -0.0000

Data Index: 907388 (The Pile)

Max Activating Token Index: 505

Text #17

Max Range: 7.6724. Min Range: -7.6724

Max Act: 6.4543. Min Act: -0.0000

Data Index: 327724 (The Pile)

Max Activating Token Index: 660

Text #18

Max Range: 7.6724. Min Range: -7.6724

Max Act: 6.4740. Min Act: -0.0000

Data Index: 513039 (The Pile)

Max Activating Token Index: 132

Text #19

Max Range: 7.6724. Min Range: -7.6724

Max Act: 6.2333. Min Act: -0.0000

Data Index: 503702 (The Pile)

Max Activating Token Index: 855