INDEX

Explanations

terms and attributes related to image formatting and dimensions

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

aspect ratio and orientation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.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

 Diſ

-0.52

groupby

-0.47

 Anſ

-0.47

queryInterface

-0.46

 Reſ

-0.43

 Jefus

-0.43

peech

-0.42

 Inſ

-0.42

 Meksiku

-0.42

 Normdatei

-0.41

POSITIVE LOGITS

aspectRatio

0.64

setVerticalGroup

0.58

sidemargin

0.49

 ratios

0.48

 ratio

0.47

idescreen

0.47

RATIO

0.45

 arşivlendi

0.44

 BoxFit

0.44

MeasureSpec

0.44

Activations Density 0.008%

terms and attributes related to image formatting and dimensions

aspect ratio and orientation

No Comments

No Known Activations