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

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

Neuron 3715 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 1.8882. Min Range: -1.8882

Max Act: 1.8882. Min Act: -0.0001

Data Index: 467140 (C4 (Web Text))

Max Activating Token Index: 849

Text #1

Max Range: 1.8882. Min Range: -1.8882

Max Act: 1.7376. Min Act: -0.0001

Data Index: 349716 (C4 (Web Text))

Max Activating Token Index: 152

Text #2

Max Range: 1.8882. Min Range: -1.8882

Max Act: 1.6721. Min Act: -0.0001

Data Index: 880831 (C4 (Web Text))

Max Activating Token Index: 745

Text #3

Max Range: 1.8882. Min Range: -1.8882

Max Act: 1.7128. Min Act: -0.0001

Data Index: 159180 (C4 (Web Text))

Max Activating Token Index: 427

Text #4

Max Range: 1.8882. Min Range: -1.8882

Max Act: 1.5809. Min Act: -0.0001

Data Index: 342102 (C4 (Web Text))

Max Activating Token Index: 497

Text #5

Max Range: 1.8882. Min Range: -1.8882

Max Act: 1.6138. Min Act: -0.0001

Data Index: 121440 (C4 (Web Text))

Max Activating Token Index: 262

Text #6

Max Range: 1.8882. Min Range: -1.8882

Max Act: 1.5217. Min Act: -0.0001

Data Index: 1543793 (Python Code)

Max Activating Token Index: 158

Text #7

Max Range: 1.8882. Min Range: -1.8882

Max Act: 1.5118. Min Act: -0.0001

Data Index: 159916 (C4 (Web Text))

Max Activating Token Index: 130

Text #8

Max Range: 1.8882. Min Range: -1.8882

Max Act: 1.4540. Min Act: -0.0001

Data Index: 1216301 (C4 (Web Text))

Max Activating Token Index: 350

Text #9

Max Range: 1.8882. Min Range: -1.8882

Max Act: 1.4588. Min Act: -0.0001

Data Index: 1179134 (C4 (Web Text))

Max Activating Token Index: 910

Text #10

Max Range: 1.8882. Min Range: -1.8882

Max Act: 1.4082. Min Act: -0.0001

Data Index: 1483204 (Python Code)

Max Activating Token Index: 897

Text #11

Max Range: 1.8882. Min Range: -1.8882

Max Act: 1.3939. Min Act: -0.0001

Data Index: 1371185 (Python Code)

Max Activating Token Index: 743

Text #12

Max Range: 1.8882. Min Range: -1.8882

Max Act: 1.4251. Min Act: -0.0001

Data Index: 38901 (C4 (Web Text))

Max Activating Token Index: 196

Text #13

Max Range: 1.8882. Min Range: -1.8882

Max Act: 1.3833. Min Act: -0.0001

Data Index: 1505972 (Python Code)

Max Activating Token Index: 168

Text #14

Max Range: 1.8882. Min Range: -1.8882

Max Act: 1.3705. Min Act: -0.0001

Data Index: 408709 (C4 (Web Text))

Max Activating Token Index: 605

Text #15

Max Range: 1.8882. Min Range: -1.8882

Max Act: 1.4019. Min Act: -0.0001

Data Index: 1559022 (Python Code)

Max Activating Token Index: 155

Text #16

Max Range: 1.8882. Min Range: -1.8882

Max Act: 1.3215. Min Act: -0.0001

Data Index: 414367 (C4 (Web Text))

Max Activating Token Index: 869

Text #17

Max Range: 1.8882. Min Range: -1.8882

Max Act: 1.3968. Min Act: -0.0001

Data Index: 1453169 (Python Code)

Max Activating Token Index: 652

Text #18

Max Range: 1.8882. Min Range: -1.8882

Max Act: 1.3030. Min Act: -0.0001

Data Index: 602140 (C4 (Web Text))

Max Activating Token Index: 303

Text #19

Max Range: 1.8882. Min Range: -1.8882

Max Act: 1.3325. Min Act: -0.0001

Data Index: 294847 (C4 (Web Text))

Max Activating Token Index: 697