Model: SoLU Model: 3 Layers, 2048 Neurons per Layer

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

Neuron 1757 in Layer 0

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-3l')

Text #0

Max Range: 0.0160. Min Range: -0.0160

Max Act: 0.0160. Min Act: -0.0006

Data Index: 175680 (C4 (Web Text))

Max Activating Token Index: 94

Text #1

Max Range: 0.0160. Min Range: -0.0160

Max Act: 0.0159. Min Act: -0.0006

Data Index: 548427 (C4 (Web Text))

Max Activating Token Index: 673

Text #2

Max Range: 0.0160. Min Range: -0.0160

Max Act: 0.0157. Min Act: -0.0005

Data Index: 71871 (C4 (Web Text))

Max Activating Token Index: 417

Text #3

Max Range: 0.0160. Min Range: -0.0160

Max Act: 0.0160. Min Act: -0.0006

Data Index: 1599038 (Python Code)

Max Activating Token Index: 914

Text #4

Max Range: 0.0160. Min Range: -0.0160

Max Act: 0.0152. Min Act: -0.0006

Data Index: 500593 (C4 (Web Text))

Max Activating Token Index: 624

Text #5

Max Range: 0.0160. Min Range: -0.0160

Max Act: 0.0153. Min Act: -0.0006

Data Index: 741311 (C4 (Web Text))

Max Activating Token Index: 43

Text #6

Max Range: 0.0160. Min Range: -0.0160

Max Act: 0.0150. Min Act: -0.0006

Data Index: 777652 (C4 (Web Text))

Max Activating Token Index: 734

Text #7

Max Range: 0.0160. Min Range: -0.0160

Max Act: 0.0148. Min Act: -0.0005

Data Index: 491635 (C4 (Web Text))

Max Activating Token Index: 381

Text #8

Max Range: 0.0160. Min Range: -0.0160

Max Act: 0.0149. Min Act: -0.0006

Data Index: 1147144 (C4 (Web Text))

Max Activating Token Index: 204

Text #9

Max Range: 0.0160. Min Range: -0.0160

Max Act: 0.0145. Min Act: -0.0005

Data Index: 1178395 (C4 (Web Text))

Max Activating Token Index: 206

Text #10

Max Range: 0.0160. Min Range: -0.0160

Max Act: 0.0143. Min Act: -0.0006

Data Index: 1583313 (Python Code)

Max Activating Token Index: 806

Text #11

Max Range: 0.0160. Min Range: -0.0160

Max Act: 0.0143. Min Act: -0.0006

Data Index: 1581029 (Python Code)

Max Activating Token Index: 306

Text #12

Max Range: 0.0160. Min Range: -0.0160

Max Act: 0.0142. Min Act: -0.0006

Data Index: 1347588 (C4 (Web Text))

Max Activating Token Index: 793

Text #13

Max Range: 0.0160. Min Range: -0.0160

Max Act: 0.0140. Min Act: -0.0006

Data Index: 1425019 (Python Code)

Max Activating Token Index: 534

Text #14

Max Range: 0.0160. Min Range: -0.0160

Max Act: 0.0136. Min Act: -0.0006

Data Index: 396641 (C4 (Web Text))

Max Activating Token Index: 984

Text #15

Max Range: 0.0160. Min Range: -0.0160

Max Act: 0.0139. Min Act: -0.0006

Data Index: 1415745 (Python Code)

Max Activating Token Index: 552

Text #16

Max Range: 0.0160. Min Range: -0.0160

Max Act: 0.0139. Min Act: -0.0006

Data Index: 92736 (C4 (Web Text))

Max Activating Token Index: 272

Text #17

Max Range: 0.0160. Min Range: -0.0160

Max Act: 0.0138. Min Act: -0.0006

Data Index: 1325208 (C4 (Web Text))

Max Activating Token Index: 470

Text #18

Max Range: 0.0160. Min Range: -0.0160

Max Act: 0.0137. Min Act: -0.0006

Data Index: 34149 (C4 (Web Text))

Max Activating Token Index: 39

Text #19

Max Range: 0.0160. Min Range: -0.0160

Max Act: 0.0138. Min Act: -0.0006

Data Index: 1620213 (Python Code)

Max Activating Token Index: 429