Model: SoLU Model: 12 Layers, 6144 Neurons per Layer

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

Neuron 1823 in Layer 7

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-12l')

Text #0

Max Range: 1.5788. Min Range: -1.5788

Max Act: 1.5788. Min Act: -0.0000

Data Index: 278334 (C4 (Web Text))

Max Activating Token Index: 897

Text #1

Max Range: 1.5788. Min Range: -1.5788

Max Act: 1.4670. Min Act: -0.0000

Data Index: 348097 (C4 (Web Text))

Max Activating Token Index: 309

Text #2

Max Range: 1.5788. Min Range: -1.5788

Max Act: 1.3726. Min Act: -0.0000

Data Index: 720467 (C4 (Web Text))

Max Activating Token Index: 347

Text #3

Max Range: 1.5788. Min Range: -1.5788

Max Act: 1.4847. Min Act: -0.0000

Data Index: 195005 (C4 (Web Text))

Max Activating Token Index: 242

Text #4

Max Range: 1.5788. Min Range: -1.5788

Max Act: 1.2314. Min Act: -0.0000

Data Index: 39860 (C4 (Web Text))

Max Activating Token Index: 436

Text #5

Max Range: 1.5788. Min Range: -1.5788

Max Act: 1.1258. Min Act: -0.0000

Data Index: 1423989 (Python Code)

Max Activating Token Index: 994

Text #6

Max Range: 1.5788. Min Range: -1.5788

Max Act: 1.2172. Min Act: -0.0000

Data Index: 376728 (C4 (Web Text))

Max Activating Token Index: 61

Text #7

Max Range: 1.5788. Min Range: -1.5788

Max Act: 1.1163. Min Act: -0.0000

Data Index: 1629184 (Python Code)

Max Activating Token Index: 141

Text #8

Max Range: 1.5788. Min Range: -1.5788

Max Act: 1.1134. Min Act: -0.0000

Data Index: 485942 (C4 (Web Text))

Max Activating Token Index: 80

Text #9

Max Range: 1.5788. Min Range: -1.5788

Max Act: 1.1153. Min Act: -0.0000

Data Index: 1556089 (Python Code)

Max Activating Token Index: 405

Text #10

Max Range: 1.5788. Min Range: -1.5788

Max Act: 1.1050. Min Act: -0.0000

Data Index: 871417 (C4 (Web Text))

Max Activating Token Index: 5

Text #11

Max Range: 1.5788. Min Range: -1.5788

Max Act: 1.0703. Min Act: -0.0000

Data Index: 1452267 (Python Code)

Max Activating Token Index: 498

Text #12

Max Range: 1.5788. Min Range: -1.5788

Max Act: 1.0668. Min Act: -0.0000

Data Index: 345622 (C4 (Web Text))

Max Activating Token Index: 54

Text #13

Max Range: 1.5788. Min Range: -1.5788

Max Act: 1.1566. Min Act: -0.0000

Data Index: 1566301 (Python Code)

Max Activating Token Index: 848

Text #14

Max Range: 1.5788. Min Range: -1.5788

Max Act: 1.1029. Min Act: -0.0000

Data Index: 1393990 (Python Code)

Max Activating Token Index: 466

Text #15

Max Range: 1.5788. Min Range: -1.5788

Max Act: 1.1582. Min Act: -0.0000

Data Index: 1217555 (C4 (Web Text))

Max Activating Token Index: 1002

Text #16

Max Range: 1.5788. Min Range: -1.5788

Max Act: 1.0555. Min Act: -0.0000

Data Index: 1525169 (Python Code)

Max Activating Token Index: 267

Text #17

Max Range: 1.5788. Min Range: -1.5788

Max Act: 1.0237. Min Act: -0.0000

Data Index: 1563818 (Python Code)

Max Activating Token Index: 646

Text #18

Max Range: 1.5788. Min Range: -1.5788

Max Act: 0.9856. Min Act: -0.0000

Data Index: 1616248 (Python Code)

Max Activating Token Index: 842

Text #19

Max Range: 1.5788. Min Range: -1.5788

Max Act: 1.0406. Min Act: -0.0000

Data Index: 1264055 (C4 (Web Text))

Max Activating Token Index: 537