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

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

Neuron 1789 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 5.3843. Min Range: -5.3843

Max Act: 5.3531. Min Act: -0.1700

Data Index: 903166 (C4 (Web Text))

Max Activating Token Index: 503

Text #1

Max Range: 5.3843. Min Range: -5.3843

Max Act: 5.3843. Min Act: -0.1700

Data Index: 1527387 (Python Code)

Max Activating Token Index: 238

Text #2

Max Range: 5.3843. Min Range: -5.3843

Max Act: 5.3218. Min Act: -0.1700

Data Index: 1433158 (Python Code)

Max Activating Token Index: 241

Text #3

Max Range: 5.3843. Min Range: -5.3843

Max Act: 5.1968. Min Act: -0.1700

Data Index: 1603999 (Python Code)

Max Activating Token Index: 261

Text #4

Max Range: 5.3843. Min Range: -5.3843

Max Act: 5.2281. Min Act: -0.1700

Data Index: 1619383 (Python Code)

Max Activating Token Index: 237

Text #5

Max Range: 5.3843. Min Range: -5.3843

Max Act: 5.2281. Min Act: -0.1700

Data Index: 1639766 (Python Code)

Max Activating Token Index: 255

Text #6

Max Range: 5.3843. Min Range: -5.3843

Max Act: 5.1968. Min Act: -0.1699

Data Index: 1548336 (Python Code)

Max Activating Token Index: 209

Text #7

Max Range: 5.3843. Min Range: -5.3843

Max Act: 5.0718. Min Act: -0.1700

Data Index: 145792 (C4 (Web Text))

Max Activating Token Index: 328

Text #8

Max Range: 5.3843. Min Range: -5.3843

Max Act: 5.0718. Min Act: -0.1700

Data Index: 1409434 (Python Code)

Max Activating Token Index: 727

Text #9

Max Range: 5.3843. Min Range: -5.3843

Max Act: 5.0718. Min Act: -0.1700

Data Index: 1428335 (Python Code)

Max Activating Token Index: 184

Text #10

Max Range: 5.3843. Min Range: -5.3843

Max Act: 5.1031. Min Act: -0.1700

Data Index: 1412439 (Python Code)

Max Activating Token Index: 415

Text #11

Max Range: 5.3843. Min Range: -5.3843

Max Act: 5.0718. Min Act: -0.1700

Data Index: 1440200 (Python Code)

Max Activating Token Index: 189

Text #12

Max Range: 5.3843. Min Range: -5.3843

Max Act: 5.0093. Min Act: -0.1700

Data Index: 1363942 (Python Code)

Max Activating Token Index: 194

Text #13

Max Range: 5.3843. Min Range: -5.3843

Max Act: 5.0718. Min Act: -0.1700

Data Index: 1445417 (Python Code)

Max Activating Token Index: 613

Text #14

Max Range: 5.3843. Min Range: -5.3843

Max Act: 5.0406. Min Act: -0.1700

Data Index: 1470226 (Python Code)

Max Activating Token Index: 203

Text #15

Max Range: 5.3843. Min Range: -5.3843

Max Act: 4.9468. Min Act: -0.1700

Data Index: 1568404 (Python Code)

Max Activating Token Index: 481

Text #16

Max Range: 5.3843. Min Range: -5.3843

Max Act: 5.0093. Min Act: -0.1700

Data Index: 1361335 (Python Code)

Max Activating Token Index: 108

Text #17

Max Range: 5.3843. Min Range: -5.3843

Max Act: 5.0093. Min Act: -0.1700

Data Index: 1013635 (C4 (Web Text))

Max Activating Token Index: 781

Text #18

Max Range: 5.3843. Min Range: -5.3843

Max Act: 5.0093. Min Act: -0.1700

Data Index: 1634474 (Python Code)

Max Activating Token Index: 415

Text #19

Max Range: 5.3843. Min Range: -5.3843

Max Act: 4.9468. Min Act: -0.1700

Data Index: 1561732 (Python Code)

Max Activating Token Index: 277