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

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

Neuron 637 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: 5.7519. Min Range: -5.7519

Max Act: 5.7519. Min Act: -0.1700

Data Index: 1339542 (C4 (Web Text))

Max Activating Token Index: 412

Text #1

Max Range: 5.7519. Min Range: -5.7519

Max Act: 5.6582. Min Act: -0.1700

Data Index: 1264335 (C4 (Web Text))

Max Activating Token Index: 985

Text #2

Max Range: 5.7519. Min Range: -5.7519

Max Act: 5.5019. Min Act: -0.1700

Data Index: 1135770 (C4 (Web Text))

Max Activating Token Index: 621

Text #3

Max Range: 5.7519. Min Range: -5.7519

Max Act: 5.3457. Min Act: -0.1700

Data Index: 453914 (C4 (Web Text))

Max Activating Token Index: 158

Text #4

Max Range: 5.7519. Min Range: -5.7519

Max Act: 5.3144. Min Act: -0.1700

Data Index: 190977 (C4 (Web Text))

Max Activating Token Index: 531

Text #5

Max Range: 5.7519. Min Range: -5.7519

Max Act: 5.3144. Min Act: -0.1700

Data Index: 135836 (C4 (Web Text))

Max Activating Token Index: 422

Text #6

Max Range: 5.7519. Min Range: -5.7519

Max Act: 5.2519. Min Act: -0.1700

Data Index: 1011074 (C4 (Web Text))

Max Activating Token Index: 950

Text #7

Max Range: 5.7519. Min Range: -5.7519

Max Act: 5.2832. Min Act: -0.1700

Data Index: 888457 (C4 (Web Text))

Max Activating Token Index: 113

Text #8

Max Range: 5.7519. Min Range: -5.7519

Max Act: 5.2519. Min Act: -0.1700

Data Index: 886530 (C4 (Web Text))

Max Activating Token Index: 150

Text #9

Max Range: 5.7519. Min Range: -5.7519

Max Act: 5.1894. Min Act: -0.1700

Data Index: 1130396 (C4 (Web Text))

Max Activating Token Index: 180

Text #10

Max Range: 5.7519. Min Range: -5.7519

Max Act: 5.1582. Min Act: -0.1700

Data Index: 365444 (C4 (Web Text))

Max Activating Token Index: 752

Text #11

Max Range: 5.7519. Min Range: -5.7519

Max Act: 5.0644. Min Act: -0.1700

Data Index: 1555963 (Python Code)

Max Activating Token Index: 912

Text #12

Max Range: 5.7519. Min Range: -5.7519

Max Act: 5.1269. Min Act: -0.1700

Data Index: 72385 (C4 (Web Text))

Max Activating Token Index: 180

Text #13

Max Range: 5.7519. Min Range: -5.7519

Max Act: 4.9707. Min Act: -0.1700

Data Index: 335799 (C4 (Web Text))

Max Activating Token Index: 186

Text #14

Max Range: 5.7519. Min Range: -5.7519

Max Act: 5.0332. Min Act: -0.1700

Data Index: 1402824 (Python Code)

Max Activating Token Index: 647

Text #15

Max Range: 5.7519. Min Range: -5.7519

Max Act: 5.0019. Min Act: -0.1700

Data Index: 1410277 (Python Code)

Max Activating Token Index: 428

Text #16

Max Range: 5.7519. Min Range: -5.7519

Max Act: 5.0019. Min Act: -0.1700

Data Index: 905466 (C4 (Web Text))

Max Activating Token Index: 440

Text #17

Max Range: 5.7519. Min Range: -5.7519

Max Act: 5.0019. Min Act: -0.1700

Data Index: 1478023 (Python Code)

Max Activating Token Index: 567

Text #18

Max Range: 5.7519. Min Range: -5.7519

Max Act: 5.0332. Min Act: -0.1700

Data Index: 1421175 (Python Code)

Max Activating Token Index: 400

Text #19

Max Range: 5.7519. Min Range: -5.7519

Max Act: 5.0019. Min Act: -0.1700

Data Index: 428668 (C4 (Web Text))

Max Activating Token Index: 141