INDEX

Explanations

mentions of fake news and media literacy

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

fake news

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

Predecesor

-0.40

HOR

-0.35

 außer

-0.34

pê

-0.33

delegate

-0.32

Aholisi

-0.32

 başladı

-0.32

HOR

-0.32

길

-0.31

デューサー

-0.31

POSITIVE LOGITS

 disinformation

0.62

 misinformation

0.60

########.

0.53

twimg

0.53

 HasFactory

0.52

 ligiloj

0.52

IsContent

0.51

RenderAtEndOf

0.49

cupertino

0.49

 برانيه

0.48

Activations Density 0.016%

mentions of fake news and media literacy

fake news

No Comments

No Known Activations