INDEX

Explanations

words related to pressing or input actions, particularly in the context of technology or media

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 12-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_12/width_16k/average_l0_130

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.12.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

signable

-0.62

enumii

-0.61

 дописавши

-0.59

lihood

-0.57

ftagPool

-0.54

astéroïdes

-0.54



-0.52

RectangleBorder

-0.51

 Plut

-0.50

 kaarangay

-0.50

POSITIVE LOGITS

 Press

1.59

 press

1.55

Press

1.45

press

1.41

 PRESS

1.38

 presses

1.23

 Pressing

1.21

PRESS

1.17

 pressed

1.13

 pressing

1.11

Activations Density 0.160%

words related to pressing or input actions, particularly in the context of technology or media

No Comments

No Known Activations