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

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

Neuron 1592 in Layer 0

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.1289. Min Range: -3.1289

Max Act: 3.1289. Min Act: -0.1700

Data Index: 1095968 (C4 (Web Text))

Max Activating Token Index: 229

Text #1

Max Range: 3.1289. Min Range: -3.1289

Max Act: 3.0974. Min Act: -0.1700

Data Index: 555471 (C4 (Web Text))

Max Activating Token Index: 86

Text #2

Max Range: 3.1289. Min Range: -3.1289

Max Act: 3.0974. Min Act: -0.1700

Data Index: 1643074 (Python Code)

Max Activating Token Index: 799

Text #3

Max Range: 3.1289. Min Range: -3.1289

Max Act: 3.0816. Min Act: -0.1700

Data Index: 303939 (C4 (Web Text))

Max Activating Token Index: 941

Text #4

Max Range: 3.1289. Min Range: -3.1289

Max Act: 3.0816. Min Act: -0.1700

Data Index: 301516 (C4 (Web Text))

Max Activating Token Index: 57

Text #5

Max Range: 3.1289. Min Range: -3.1289

Max Act: 3.0343. Min Act: -0.1700

Data Index: 814574 (C4 (Web Text))

Max Activating Token Index: 32

Text #6

Max Range: 3.1289. Min Range: -3.1289

Max Act: 3.0501. Min Act: -0.1700

Data Index: 937391 (C4 (Web Text))

Max Activating Token Index: 990

Text #7

Max Range: 3.1289. Min Range: -3.1289

Max Act: 3.0501. Min Act: -0.1700

Data Index: 1303026 (C4 (Web Text))

Max Activating Token Index: 1010

Text #8

Max Range: 3.1289. Min Range: -3.1289

Max Act: 3.0185. Min Act: -0.1700

Data Index: 349372 (C4 (Web Text))

Max Activating Token Index: 788

Text #9

Max Range: 3.1289. Min Range: -3.1289

Max Act: 3.0343. Min Act: -0.1700

Data Index: 257471 (C4 (Web Text))

Max Activating Token Index: 794

Text #10

Max Range: 3.1289. Min Range: -3.1289

Max Act: 3.0343. Min Act: -0.1700

Data Index: 10198 (C4 (Web Text))

Max Activating Token Index: 369

Text #11

Max Range: 3.1289. Min Range: -3.1289

Max Act: 3.0343. Min Act: -0.1700

Data Index: 746759 (C4 (Web Text))

Max Activating Token Index: 973

Text #12

Max Range: 3.1289. Min Range: -3.1289

Max Act: 3.0343. Min Act: -0.1700

Data Index: 493200 (C4 (Web Text))

Max Activating Token Index: 905

Text #13

Max Range: 3.1289. Min Range: -3.1289

Max Act: 3.0343. Min Act: -0.1700

Data Index: 1517330 (Python Code)

Max Activating Token Index: 412

Text #14

Max Range: 3.1289. Min Range: -3.1289

Max Act: 3.0185. Min Act: -0.1700

Data Index: 17521 (C4 (Web Text))

Max Activating Token Index: 829

Text #15

Max Range: 3.1289. Min Range: -3.1289

Max Act: 3.0185. Min Act: -0.1700

Data Index: 629746 (C4 (Web Text))

Max Activating Token Index: 568

Text #16

Max Range: 3.1289. Min Range: -3.1289

Max Act: 3.0027. Min Act: -0.1700

Data Index: 584282 (C4 (Web Text))

Max Activating Token Index: 827

Text #17

Max Range: 3.1289. Min Range: -3.1289

Max Act: 3.0027. Min Act: -0.1700

Data Index: 456235 (C4 (Web Text))

Max Activating Token Index: 289

Text #18

Max Range: 3.1289. Min Range: -3.1289

Max Act: 3.0185. Min Act: -0.1700

Data Index: 157271 (C4 (Web Text))

Max Activating Token Index: 850

Text #19

Max Range: 3.1289. Min Range: -3.1289

Max Act: 3.0027. Min Act: -0.1700

Data Index: 84284 (C4 (Web Text))

Max Activating Token Index: 570