INDEX

Explanations

phrases indicating prominence or recognition

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

urlpatterns

-0.70

präche

-0.67

abetes

-0.63

emptory

-0.61

выше

-0.56

 guaranty

-0.53

iastes

-0.53

ברס

-0.53

 unangemessene

-0.52

 Extraordinary

-0.52

POSITIVE LOGITS

antMatchers

0.65

asList

0.58

 primarily

0.52

ByUrl

0.52

gså

0.50

bibitem

0.49

})));

0.48

AppCompatTheme

0.47

+#+#

0.47

 Приступљено

0.47

Activations Density 0.358%

phrases indicating prominence or recognition

No Comments

No Known Activations