Model: SoLU Model: 4 Layers, 2048 Neurons per Layer

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

Neuron 480 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-4l')

Text #0

Max Range: 0.0105. Min Range: -0.0105

Max Act: 0.0105. Min Act: -0.0006

Data Index: 369128 (C4 (Web Text))

Max Activating Token Index: 465

Text #1

Max Range: 0.0105. Min Range: -0.0105

Max Act: 0.0099. Min Act: -0.0006

Data Index: 52545 (C4 (Web Text))

Max Activating Token Index: 906

Text #2

Max Range: 0.0105. Min Range: -0.0105

Max Act: 0.0099. Min Act: -0.0006

Data Index: 1072309 (C4 (Web Text))

Max Activating Token Index: 630

Text #3

Max Range: 0.0105. Min Range: -0.0105

Max Act: 0.0105. Min Act: -0.0006

Data Index: 1255683 (C4 (Web Text))

Max Activating Token Index: 870

Text #4

Max Range: 0.0105. Min Range: -0.0105

Max Act: 0.0084. Min Act: -0.0006

Data Index: 105270 (C4 (Web Text))

Max Activating Token Index: 721

Text #5

Max Range: 0.0105. Min Range: -0.0105

Max Act: 0.0088. Min Act: -0.0006

Data Index: 610493 (C4 (Web Text))

Max Activating Token Index: 920

Text #6

Max Range: 0.0105. Min Range: -0.0105

Max Act: 0.0085. Min Act: -0.0006

Data Index: 1005625 (C4 (Web Text))

Max Activating Token Index: 705

Text #7

Max Range: 0.0105. Min Range: -0.0105

Max Act: 0.0081. Min Act: -0.0006

Data Index: 74419 (C4 (Web Text))

Max Activating Token Index: 703

Text #8

Max Range: 0.0105. Min Range: -0.0105

Max Act: 0.0087. Min Act: -0.0006

Data Index: 568206 (C4 (Web Text))

Max Activating Token Index: 530

Text #9

Max Range: 0.0105. Min Range: -0.0105

Max Act: 0.0085. Min Act: -0.0006

Data Index: 124790 (C4 (Web Text))

Max Activating Token Index: 375

Text #10

Max Range: 0.0105. Min Range: -0.0105

Max Act: 0.0082. Min Act: -0.0006

Data Index: 88681 (C4 (Web Text))

Max Activating Token Index: 396

Text #11

Max Range: 0.0105. Min Range: -0.0105

Max Act: 0.0084. Min Act: -0.0006

Data Index: 21088 (C4 (Web Text))

Max Activating Token Index: 585

Text #12

Max Range: 0.0105. Min Range: -0.0105

Max Act: 0.0086. Min Act: -0.0006

Data Index: 1013765 (C4 (Web Text))

Max Activating Token Index: 429

Text #13

Max Range: 0.0105. Min Range: -0.0105

Max Act: 0.0081. Min Act: -0.0006

Data Index: 1169808 (C4 (Web Text))

Max Activating Token Index: 923

Text #14

Max Range: 0.0105. Min Range: -0.0105

Max Act: 0.0080. Min Act: -0.0006

Data Index: 881242 (C4 (Web Text))

Max Activating Token Index: 883

Text #15

Max Range: 0.0105. Min Range: -0.0105

Max Act: 0.0076. Min Act: -0.0006

Data Index: 762417 (C4 (Web Text))

Max Activating Token Index: 923

Text #16

Max Range: 0.0105. Min Range: -0.0105

Max Act: 0.0078. Min Act: -0.0006

Data Index: 1527630 (Python Code)

Max Activating Token Index: 394

Text #17

Max Range: 0.0105. Min Range: -0.0105

Max Act: 0.0078. Min Act: -0.0006

Data Index: 1052004 (C4 (Web Text))

Max Activating Token Index: 643

Text #18

Max Range: 0.0105. Min Range: -0.0105

Max Act: 0.0075. Min Act: -0.0006

Data Index: 104454 (C4 (Web Text))

Max Activating Token Index: 835

Text #19

Max Range: 0.0105. Min Range: -0.0105

Max Act: 0.0083. Min Act: -0.0006

Data Index: 661642 (C4 (Web Text))

Max Activating Token Index: 414