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

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

Neuron 3095 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 1.4937. Min Range: -1.4937

Max Act: 1.3944. Min Act: -0.0001

Data Index: 1053267 (C4 (Web Text))

Max Activating Token Index: 159

Text #1

Max Range: 1.4937. Min Range: -1.4937

Max Act: 1.4496. Min Act: -0.0001

Data Index: 48691 (C4 (Web Text))

Max Activating Token Index: 552

Text #2

Max Range: 1.4937. Min Range: -1.4937

Max Act: 1.4937. Min Act: -0.0001

Data Index: 837225 (C4 (Web Text))

Max Activating Token Index: 394

Text #3

Max Range: 1.4937. Min Range: -1.4937

Max Act: 1.3016. Min Act: -0.0001

Data Index: 1053890 (C4 (Web Text))

Max Activating Token Index: 658

Text #4

Max Range: 1.4937. Min Range: -1.4937

Max Act: 1.2337. Min Act: -0.0001

Data Index: 1337019 (C4 (Web Text))

Max Activating Token Index: 740

Text #5

Max Range: 1.4937. Min Range: -1.4937

Max Act: 1.2671. Min Act: -0.0001

Data Index: 1339146 (C4 (Web Text))

Max Activating Token Index: 638

Text #6

Max Range: 1.4937. Min Range: -1.4937

Max Act: 1.2818. Min Act: -0.0001

Data Index: 1316786 (C4 (Web Text))

Max Activating Token Index: 261

Text #7

Max Range: 1.4937. Min Range: -1.4937

Max Act: 1.1882. Min Act: -0.0001

Data Index: 890054 (C4 (Web Text))

Max Activating Token Index: 4

Text #8

Max Range: 1.4937. Min Range: -1.4937

Max Act: 1.2959. Min Act: -0.0001

Data Index: 609083 (C4 (Web Text))

Max Activating Token Index: 682

Text #9

Max Range: 1.4937. Min Range: -1.4937

Max Act: 1.1267. Min Act: -0.0001

Data Index: 893573 (C4 (Web Text))

Max Activating Token Index: 19

Text #10

Max Range: 1.4937. Min Range: -1.4937

Max Act: 1.2377. Min Act: -0.0001

Data Index: 13983 (C4 (Web Text))

Max Activating Token Index: 256

Text #11

Max Range: 1.4937. Min Range: -1.4937

Max Act: 1.2454. Min Act: -0.0001

Data Index: 234933 (C4 (Web Text))

Max Activating Token Index: 56

Text #12

Max Range: 1.4937. Min Range: -1.4937

Max Act: 1.2060. Min Act: -0.0001

Data Index: 485325 (C4 (Web Text))

Max Activating Token Index: 806

Text #13

Max Range: 1.4937. Min Range: -1.4937

Max Act: 1.2533. Min Act: -0.0001

Data Index: 13764 (C4 (Web Text))

Max Activating Token Index: 812

Text #14

Max Range: 1.4937. Min Range: -1.4937

Max Act: 1.1376. Min Act: -0.0001

Data Index: 720166 (C4 (Web Text))

Max Activating Token Index: 191

Text #15

Max Range: 1.4937. Min Range: -1.4937

Max Act: 1.1954. Min Act: -0.0001

Data Index: 1125850 (C4 (Web Text))

Max Activating Token Index: 144

Text #16

Max Range: 1.4937. Min Range: -1.4937

Max Act: 1.0842. Min Act: -0.0001

Data Index: 939840 (C4 (Web Text))

Max Activating Token Index: 505

Text #17

Max Range: 1.4937. Min Range: -1.4937

Max Act: 1.0718. Min Act: -0.0001

Data Index: 391563 (C4 (Web Text))

Max Activating Token Index: 557

Text #18

Max Range: 1.4937. Min Range: -1.4937

Max Act: 1.1174. Min Act: -0.0001

Data Index: 1586522 (Python Code)

Max Activating Token Index: 236

Text #19

Max Range: 1.4937. Min Range: -1.4937

Max Act: 1.0233. Min Act: -0.0001

Data Index: 478080 (C4 (Web Text))

Max Activating Token Index: 50