INDEX

Explanations

reference to people, particularly pronouns like "he" and "she."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 4-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_4/width_16k/average_l0_124

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.4.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

Чыганаклар

-0.64

RegressionTest

-0.62

}*/

-0.62

ுக்கு

-0.60

 autorytatywna

-0.58

綜合報導

-0.58

expandindo

-0.56

Aholisi

-0.55

<bos>

-0.54

 الرياضيه

-0.54

POSITIVE LOGITS

HtmlAttribute

0.76

VersionUID

0.70

ValueStyle

0.67

neurial

0.62

WebVitals

0.60

RenderAtEndOf

0.58

 AttributeSet

0.58

IsContent

0.57

setVerticalGroup

0.57

MigrationBuilder

0.55

Activations Density 0.070%

reference to people, particularly pronouns like "he" and "she."

No Comments

No Known Activations