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

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

Neuron 309 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.3930. Min Range: -3.3930

Max Act: 3.3930. Min Act: -0.1700

Data Index: 266048 (C4 (Web Text))

Max Activating Token Index: 774

Text #1

Max Range: 3.3930. Min Range: -3.3930

Max Act: 3.3302. Min Act: -0.1700

Data Index: 383491 (C4 (Web Text))

Max Activating Token Index: 859

Text #2

Max Range: 3.3930. Min Range: -3.3930

Max Act: 3.3616. Min Act: -0.1700

Data Index: 292051 (C4 (Web Text))

Max Activating Token Index: 221

Text #3

Max Range: 3.3930. Min Range: -3.3930

Max Act: 3.2988. Min Act: -0.1700

Data Index: 173715 (C4 (Web Text))

Max Activating Token Index: 567

Text #4

Max Range: 3.3930. Min Range: -3.3930

Max Act: 3.2988. Min Act: -0.1700

Data Index: 958794 (C4 (Web Text))

Max Activating Token Index: 353

Text #5

Max Range: 3.3930. Min Range: -3.3930

Max Act: 3.2360. Min Act: -0.1700

Data Index: 1404867 (Python Code)

Max Activating Token Index: 241

Text #6

Max Range: 3.3930. Min Range: -3.3930

Max Act: 3.2988. Min Act: -0.1700

Data Index: 1560594 (Python Code)

Max Activating Token Index: 805

Text #7

Max Range: 3.3930. Min Range: -3.3930

Max Act: 3.2674. Min Act: -0.1700

Data Index: 439733 (C4 (Web Text))

Max Activating Token Index: 978

Text #8

Max Range: 3.3930. Min Range: -3.3930

Max Act: 3.2360. Min Act: -0.1700

Data Index: 1545940 (Python Code)

Max Activating Token Index: 739

Text #9

Max Range: 3.3930. Min Range: -3.3930

Max Act: 3.2360. Min Act: -0.1700

Data Index: 1386894 (Python Code)

Max Activating Token Index: 109

Text #10

Max Range: 3.3930. Min Range: -3.3930

Max Act: 3.2045. Min Act: -0.1700

Data Index: 263254 (C4 (Web Text))

Max Activating Token Index: 486

Text #11

Max Range: 3.3930. Min Range: -3.3930

Max Act: 3.2360. Min Act: -0.1700

Data Index: 777231 (C4 (Web Text))

Max Activating Token Index: 907

Text #12

Max Range: 3.3930. Min Range: -3.3930

Max Act: 3.1730. Min Act: -0.1700

Data Index: 1579217 (Python Code)

Max Activating Token Index: 730

Text #13

Max Range: 3.3930. Min Range: -3.3930

Max Act: 3.2045. Min Act: -0.1700

Data Index: 1491445 (Python Code)

Max Activating Token Index: 523

Text #14

Max Range: 3.3930. Min Range: -3.3930

Max Act: 3.2045. Min Act: -0.1700

Data Index: 1380277 (Python Code)

Max Activating Token Index: 629

Text #15

Max Range: 3.3930. Min Range: -3.3930

Max Act: 3.2045. Min Act: -0.1700

Data Index: 887031 (C4 (Web Text))

Max Activating Token Index: 19

Text #16

Max Range: 3.3930. Min Range: -3.3930

Max Act: 3.1730. Min Act: -0.1700

Data Index: 376437 (C4 (Web Text))

Max Activating Token Index: 504

Text #17

Max Range: 3.3930. Min Range: -3.3930

Max Act: 3.1415. Min Act: -0.1700

Data Index: 1215939 (C4 (Web Text))

Max Activating Token Index: 92

Text #18

Max Range: 3.3930. Min Range: -3.3930

Max Act: 3.2045. Min Act: -0.1700

Data Index: 853383 (C4 (Web Text))

Max Activating Token Index: 731

Text #19

Max Range: 3.3930. Min Range: -3.3930

Max Act: 3.1730. Min Act: -0.1700

Data Index: 1610581 (Python Code)

Max Activating Token Index: 212