INDEX

Explanations

actions related to violence and conflict

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

war and conflict

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

MLLoader

-0.58

 noDo

-0.44

 desglose

-0.44

 '\\;'

-0.43

 pageContext

-0.43

 StatefulWidget

-0.43

ukone

-0.42

writeFieldEnd

-0.42

 Biôgrafia

-0.42

hidupan

-0.41

POSITIVE LOGITS

KURZBESCHREIBUNG

0.48

fly

0.42

 seize

0.41

fly

0.40

 chase

0.39

Chase

0.37

fry

0.36

hos

0.36

ra

0.36

Fry

0.36

Activations Density 0.090%

actions related to violence and conflict

war and conflict

No Comments

No Known Activations