INDEX

Explanations

references to structures and nomenclature in technical or programming contexts

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

numbers and labels

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

 FontFamily

-0.42

setPower

-0.40

XmlAccessorType

-0.40

ItemLayout

-0.40

曖昧さ回避

-0.39

èdia

-0.39

Litt

-0.35

ヅ

-0.35

 emot

-0.35

 Litt

-0.35

POSITIVE LOGITS

 snippetHide

0.54

 <=",

0.51

Espèce

0.50

ChildScrollView

0.48

ReusableCell

0.48

 وتسجيلات

0.47

Viitteet

0.47

makeConstraints

0.47

UrlResolution

0.45

GOTREF

0.44

Activations Density 0.204%

references to structures and nomenclature in technical or programming contexts

numbers and labels

No Comments

No Known Activations