INDEX

Explanations

geographical references and national identities in relation to Australia and the US

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

countries and regions

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

Personendaten

-0.57

parsedMessage

-0.50

babkan

-0.48

 saites

-0.45

jsonwebtoken

-0.45

ImageContext

-0.45

EDEFAULT

-0.44

GEBURTSDATUM

-0.44

pixabay

-0.44

 dAtA

-0.44

POSITIVE LOGITS

 domestic

0.48

domestic

0.47

 domestiques

0.44

local

0.42

 domestically

0.41

domés

0.40

internal

0.39

Domestic

0.38

lusconi

0.37

 domestique

0.37

Activations Density 0.018%

geographical references and national identities in relation to Australia and the US

countries and regions

No Comments

No Known Activations