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

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

Neuron 378 in Layer 5

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 2.4576. Min Range: -2.4576

Max Act: 2.4576. Min Act: -0.0000

Data Index: 1146493 (C4 (Web Text))

Max Activating Token Index: 822

Text #1

Max Range: 2.4576. Min Range: -2.4576

Max Act: 2.0061. Min Act: -0.0000

Data Index: 1054174 (C4 (Web Text))

Max Activating Token Index: 366

Text #2

Max Range: 2.4576. Min Range: -2.4576

Max Act: 1.9788. Min Act: -0.0000

Data Index: 954680 (C4 (Web Text))

Max Activating Token Index: 942

Text #3

Max Range: 2.4576. Min Range: -2.4576

Max Act: 1.8363. Min Act: -0.0000

Data Index: 1302019 (C4 (Web Text))

Max Activating Token Index: 292

Text #4

Max Range: 2.4576. Min Range: -2.4576

Max Act: 1.8956. Min Act: -0.0000

Data Index: 300584 (C4 (Web Text))

Max Activating Token Index: 279

Text #5

Max Range: 2.4576. Min Range: -2.4576

Max Act: 1.9096. Min Act: -0.0000

Data Index: 824339 (C4 (Web Text))

Max Activating Token Index: 173

Text #6

Max Range: 2.4576. Min Range: -2.4576

Max Act: 1.8361. Min Act: -0.0000

Data Index: 1578759 (Python Code)

Max Activating Token Index: 172

Text #7

Max Range: 2.4576. Min Range: -2.4576

Max Act: 1.7842. Min Act: -0.0000

Data Index: 847318 (C4 (Web Text))

Max Activating Token Index: 524

Text #8

Max Range: 2.4576. Min Range: -2.4576

Max Act: 1.8035. Min Act: -0.0000

Data Index: 1374408 (Python Code)

Max Activating Token Index: 69

Text #9

Max Range: 2.4576. Min Range: -2.4576

Max Act: 1.7310. Min Act: -0.0000

Data Index: 1473288 (Python Code)

Max Activating Token Index: 38

Text #10

Max Range: 2.4576. Min Range: -2.4576

Max Act: 1.7138. Min Act: -0.0000

Data Index: 436531 (C4 (Web Text))

Max Activating Token Index: 714

Text #11

Max Range: 2.4576. Min Range: -2.4576

Max Act: 1.6131. Min Act: -0.0000

Data Index: 742212 (C4 (Web Text))

Max Activating Token Index: 591

Text #12

Max Range: 2.4576. Min Range: -2.4576

Max Act: 1.7500. Min Act: -0.0000

Data Index: 1443370 (Python Code)

Max Activating Token Index: 87

Text #13

Max Range: 2.4576. Min Range: -2.4576

Max Act: 1.7014. Min Act: -0.0000

Data Index: 1382540 (Python Code)

Max Activating Token Index: 61

Text #14

Max Range: 2.4576. Min Range: -2.4576

Max Act: 1.5653. Min Act: -0.0000

Data Index: 1161013 (C4 (Web Text))

Max Activating Token Index: 250

Text #15

Max Range: 2.4576. Min Range: -2.4576

Max Act: 1.6441. Min Act: -0.0000

Data Index: 1582636 (Python Code)

Max Activating Token Index: 545

Text #16

Max Range: 2.4576. Min Range: -2.4576

Max Act: 1.7300. Min Act: -0.0000

Data Index: 1066329 (C4 (Web Text))

Max Activating Token Index: 241

Text #17

Max Range: 2.4576. Min Range: -2.4576

Max Act: 1.6467. Min Act: -0.0000

Data Index: 1330789 (C4 (Web Text))

Max Activating Token Index: 270

Text #18

Max Range: 2.4576. Min Range: -2.4576

Max Act: 1.6358. Min Act: -0.0000

Data Index: 14992 (C4 (Web Text))

Max Activating Token Index: 750

Text #19

Max Range: 2.4576. Min Range: -2.4576

Max Act: 1.8067. Min Act: -0.0000

Data Index: 647900 (C4 (Web Text))

Max Activating Token Index: 840