INDEX

Explanations

references to news and journalism

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

"News" or news-related terms

np_acts-logits-general · gemini-2.0-flash

news broadcasting and updates

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.18.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

aktery

-0.50

NKC

-0.49

yeb

-0.47

tente

-0.47

 بندی

-0.46

posium

-0.46

 صوتيه

-0.46

zygous

-0.45

dering

-0.45

غة

-0.45

POSITIVE LOGITS

 NEWS

0.89

feed

0.83

 news

0.78

ValueStyle

0.76

 News

0.73

NEWS

0.72

room

0.71

reel

0.71

news

0.71

flash

0.70

Activations Density 0.062%

references to news and journalism

"News" or news-related terms

news broadcasting and updates

No Comments

No Known Activations

references to news and journalism

"News" or news-related terms

news broadcasting and updates

No Comments

No Known Activations