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

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

Neuron 671 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 5.1027. Min Range: -5.1027

Max Act: 5.1027. Min Act: -0.1700

Data Index: 337323 (C4 (Web Text))

Max Activating Token Index: 71

Text #1

Max Range: 5.1027. Min Range: -5.1027

Max Act: 5.0402. Min Act: -0.1700

Data Index: 564444 (C4 (Web Text))

Max Activating Token Index: 439

Text #2

Max Range: 5.1027. Min Range: -5.1027

Max Act: 5.0089. Min Act: -0.1700

Data Index: 633500 (C4 (Web Text))

Max Activating Token Index: 872

Text #3

Max Range: 5.1027. Min Range: -5.1027

Max Act: 4.9464. Min Act: -0.1700

Data Index: 379844 (C4 (Web Text))

Max Activating Token Index: 310

Text #4

Max Range: 5.1027. Min Range: -5.1027

Max Act: 4.8839. Min Act: -0.1700

Data Index: 612886 (C4 (Web Text))

Max Activating Token Index: 34

Text #5

Max Range: 5.1027. Min Range: -5.1027

Max Act: 4.8527. Min Act: -0.1700

Data Index: 1242725 (C4 (Web Text))

Max Activating Token Index: 484

Text #6

Max Range: 5.1027. Min Range: -5.1027

Max Act: 4.8527. Min Act: -0.1700

Data Index: 964957 (C4 (Web Text))

Max Activating Token Index: 524

Text #7

Max Range: 5.1027. Min Range: -5.1027

Max Act: 4.8214. Min Act: -0.1700

Data Index: 327291 (C4 (Web Text))

Max Activating Token Index: 807

Text #8

Max Range: 5.1027. Min Range: -5.1027

Max Act: 4.8527. Min Act: -0.1700

Data Index: 51648 (C4 (Web Text))

Max Activating Token Index: 203

Text #9

Max Range: 5.1027. Min Range: -5.1027

Max Act: 4.8214. Min Act: -0.1700

Data Index: 228760 (C4 (Web Text))

Max Activating Token Index: 105

Text #10

Max Range: 5.1027. Min Range: -5.1027

Max Act: 4.8214. Min Act: -0.1700

Data Index: 880367 (C4 (Web Text))

Max Activating Token Index: 577

Text #11

Max Range: 5.1027. Min Range: -5.1027

Max Act: 4.8527. Min Act: -0.1700

Data Index: 165747 (C4 (Web Text))

Max Activating Token Index: 115

Text #12

Max Range: 5.1027. Min Range: -5.1027

Max Act: 4.8214. Min Act: -0.1700

Data Index: 1163912 (C4 (Web Text))

Max Activating Token Index: 211

Text #13

Max Range: 5.1027. Min Range: -5.1027

Max Act: 4.8214. Min Act: -0.1700

Data Index: 333281 (C4 (Web Text))

Max Activating Token Index: 238

Text #14

Max Range: 5.1027. Min Range: -5.1027

Max Act: 4.7902. Min Act: -0.1700

Data Index: 293959 (C4 (Web Text))

Max Activating Token Index: 419

Text #15

Max Range: 5.1027. Min Range: -5.1027

Max Act: 4.7589. Min Act: -0.1700

Data Index: 264640 (C4 (Web Text))

Max Activating Token Index: 12

Text #16

Max Range: 5.1027. Min Range: -5.1027

Max Act: 4.7589. Min Act: -0.1700

Data Index: 112646 (C4 (Web Text))

Max Activating Token Index: 983

Text #17

Max Range: 5.1027. Min Range: -5.1027

Max Act: 4.7902. Min Act: -0.1700

Data Index: 347060 (C4 (Web Text))

Max Activating Token Index: 414

Text #18

Max Range: 5.1027. Min Range: -5.1027

Max Act: 4.7902. Min Act: -0.1700

Data Index: 1059614 (C4 (Web Text))

Max Activating Token Index: 408

Text #19

Max Range: 5.1027. Min Range: -5.1027

Max Act: 4.7277. Min Act: -0.1700

Data Index: 8781 (C4 (Web Text))

Max Activating Token Index: 698