INDEX

Explanations

the word "following."

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-pt-res/layer_2/width_16k/average_l0_67

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.2.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

 Redewendung

-0.42

étape

-0.42

Shame

-0.37

 silvestre

-0.36

top

-0.36

 صح

-0.36

arr

-0.35

 descente

-0.35

Kanpo

-0.34

 navideños

-0.33

POSITIVE LOGITS

MLLoader

0.70

__':

0.69

 يتيمه

0.67

IUrlHelper

0.67

__':

0.66

__":

0.66

__":

0.66

DockStyle

0.62

argout

0.60

 wireType

0.57

Activations Density 0.196%

the word "following."

No Comments

No Known Activations