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

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

Neuron 995 in Layer 7

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.9177. Min Range: -0.9177

Max Act: 0.8782. Min Act: -0.0000

Data Index: 214755 (C4 (Web Text))

Max Activating Token Index: 638

Text #1

Max Range: 0.9177. Min Range: -0.9177

Max Act: 0.7918. Min Act: -0.0000

Data Index: 1010133 (C4 (Web Text))

Max Activating Token Index: 535

Text #2

Max Range: 0.9177. Min Range: -0.9177

Max Act: 0.9177. Min Act: -0.0000

Data Index: 1157200 (C4 (Web Text))

Max Activating Token Index: 688

Text #3

Max Range: 0.9177. Min Range: -0.9177

Max Act: 0.8303. Min Act: -0.0000

Data Index: 499804 (C4 (Web Text))

Max Activating Token Index: 249

Text #4

Max Range: 0.9177. Min Range: -0.9177

Max Act: 0.7337. Min Act: -0.0000

Data Index: 288519 (C4 (Web Text))

Max Activating Token Index: 261

Text #5

Max Range: 0.9177. Min Range: -0.9177

Max Act: 0.6916. Min Act: -0.0000

Data Index: 603922 (C4 (Web Text))

Max Activating Token Index: 274

Text #6

Max Range: 0.9177. Min Range: -0.9177

Max Act: 0.7126. Min Act: -0.0000

Data Index: 483806 (C4 (Web Text))

Max Activating Token Index: 327

Text #7

Max Range: 0.9177. Min Range: -0.9177

Max Act: 0.7114. Min Act: -0.0001

Data Index: 1614683 (Python Code)

Max Activating Token Index: 325

Text #8

Max Range: 0.9177. Min Range: -0.9177

Max Act: 0.7549. Min Act: -0.0001

Data Index: 1628658 (Python Code)

Max Activating Token Index: 503

Text #9

Max Range: 0.9177. Min Range: -0.9177

Max Act: 0.6944. Min Act: -0.0000

Data Index: 830700 (C4 (Web Text))

Max Activating Token Index: 389

Text #10

Max Range: 0.9177. Min Range: -0.9177

Max Act: 0.7147. Min Act: -0.0000

Data Index: 595815 (C4 (Web Text))

Max Activating Token Index: 569

Text #11

Max Range: 0.9177. Min Range: -0.9177

Max Act: 0.7012. Min Act: -0.0001

Data Index: 1467121 (Python Code)

Max Activating Token Index: 52

Text #12

Max Range: 0.9177. Min Range: -0.9177

Max Act: 0.7663. Min Act: -0.0001

Data Index: 1309182 (C4 (Web Text))

Max Activating Token Index: 586

Text #13

Max Range: 0.9177. Min Range: -0.9177

Max Act: 0.6681. Min Act: -0.0000

Data Index: 106576 (C4 (Web Text))

Max Activating Token Index: 204

Text #14

Max Range: 0.9177. Min Range: -0.9177

Max Act: 0.6529. Min Act: -0.0001

Data Index: 1143251 (C4 (Web Text))

Max Activating Token Index: 951

Text #15

Max Range: 0.9177. Min Range: -0.9177

Max Act: 0.6441. Min Act: -0.0000

Data Index: 726630 (C4 (Web Text))

Max Activating Token Index: 692

Text #16

Max Range: 0.9177. Min Range: -0.9177

Max Act: 0.7588. Min Act: -0.0001

Data Index: 598567 (C4 (Web Text))

Max Activating Token Index: 560

Text #17

Max Range: 0.9177. Min Range: -0.9177

Max Act: 0.6562. Min Act: -0.0001

Data Index: 1424545 (Python Code)

Max Activating Token Index: 413

Text #18

Max Range: 0.9177. Min Range: -0.9177

Max Act: 0.6725. Min Act: -0.0000

Data Index: 116014 (C4 (Web Text))

Max Activating Token Index: 658

Text #19

Max Range: 0.9177. Min Range: -0.9177

Max Act: 0.6270. Min Act: -0.0001

Data Index: 1463520 (Python Code)

Max Activating Token Index: 208