INDEX

Explanations

instances of images in the text

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

img followed by component names

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

ölkerung

-0.65

icitis

-0.58

 noDo

-0.58

ScopeManager

-0.57

italia

-0.57

Anhalt

-0.54

enumii

-0.53

tonsoft

-0.52

 Tartar

-0.52

cipais

-0.51

POSITIVE LOGITS

img

1.60

 image

0.57

 imgs

0.55

Img

0.48

cmd

0.48

IMG

0.47

txt

0.47

 immagin

0.47

 Image

0.46

 images

0.44

Activations Density 0.003%

instances of images in the text

img followed by component names

No Comments

No Known Activations