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

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

Neuron 4435 in Layer 6

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 2.8329. Min Range: -2.8329

Max Act: 2.8329. Min Act: -0.0000

Data Index: 340937 (C4 (Web Text))

Max Activating Token Index: 451

Text #1

Max Range: 2.8329. Min Range: -2.8329

Max Act: 2.7470. Min Act: -0.0000

Data Index: 383804 (C4 (Web Text))

Max Activating Token Index: 248

Text #2

Max Range: 2.8329. Min Range: -2.8329

Max Act: 2.1725. Min Act: -0.0000

Data Index: 826878 (C4 (Web Text))

Max Activating Token Index: 291

Text #3

Max Range: 2.8329. Min Range: -2.8329

Max Act: 2.3618. Min Act: -0.0000

Data Index: 667836 (C4 (Web Text))

Max Activating Token Index: 773

Text #4

Max Range: 2.8329. Min Range: -2.8329

Max Act: 2.2546. Min Act: -0.0000

Data Index: 1650342 (Python Code)

Max Activating Token Index: 274

Text #5

Max Range: 2.8329. Min Range: -2.8329

Max Act: 2.1124. Min Act: -0.0000

Data Index: 83774 (C4 (Web Text))

Max Activating Token Index: 200

Text #6

Max Range: 2.8329. Min Range: -2.8329

Max Act: 2.0568. Min Act: -0.0000

Data Index: 644501 (C4 (Web Text))

Max Activating Token Index: 801

Text #7

Max Range: 2.8329. Min Range: -2.8329

Max Act: 1.9257. Min Act: -0.0000

Data Index: 1026054 (C4 (Web Text))

Max Activating Token Index: 162

Text #8

Max Range: 2.8329. Min Range: -2.8329

Max Act: 1.8994. Min Act: -0.0000

Data Index: 904995 (C4 (Web Text))

Max Activating Token Index: 856

Text #9

Max Range: 2.8329. Min Range: -2.8329

Max Act: 1.9849. Min Act: -0.0000

Data Index: 1466440 (Python Code)

Max Activating Token Index: 550

Text #10

Max Range: 2.8329. Min Range: -2.8329

Max Act: 1.9698. Min Act: -0.0000

Data Index: 755215 (C4 (Web Text))

Max Activating Token Index: 464

Text #11

Max Range: 2.8329. Min Range: -2.8329

Max Act: 2.0804. Min Act: -0.0000

Data Index: 341696 (C4 (Web Text))

Max Activating Token Index: 621

Text #12

Max Range: 2.8329. Min Range: -2.8329

Max Act: 1.7272. Min Act: -0.0000

Data Index: 1386999 (Python Code)

Max Activating Token Index: 267

Text #13

Max Range: 2.8329. Min Range: -2.8329

Max Act: 1.9439. Min Act: -0.0000

Data Index: 1473151 (Python Code)

Max Activating Token Index: 246

Text #14

Max Range: 2.8329. Min Range: -2.8329

Max Act: 1.9264. Min Act: -0.0000

Data Index: 1516588 (Python Code)

Max Activating Token Index: 274

Text #15

Max Range: 2.8329. Min Range: -2.8329

Max Act: 1.9126. Min Act: -0.0000

Data Index: 855003 (C4 (Web Text))

Max Activating Token Index: 137

Text #16

Max Range: 2.8329. Min Range: -2.8329

Max Act: 2.0139. Min Act: -0.0000

Data Index: 1385868 (Python Code)

Max Activating Token Index: 746

Text #17

Max Range: 2.8329. Min Range: -2.8329

Max Act: 1.7995. Min Act: -0.0000

Data Index: 1381951 (Python Code)

Max Activating Token Index: 242

Text #18

Max Range: 2.8329. Min Range: -2.8329

Max Act: 1.9961. Min Act: -0.0000

Data Index: 1650112 (Python Code)

Max Activating Token Index: 984

Text #19

Max Range: 2.8329. Min Range: -2.8329

Max Act: 1.8910. Min Act: -0.0000

Data Index: 1417516 (Python Code)

Max Activating Token Index: 430