INDEX

Explanations

references to race, specifically the term "white" and its variations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 1-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_1/width_16k/average_l0_69

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.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

 Reverso

-0.44

__*/

-0.42

琥

-0.42

agles

-0.41

 zulassen

-0.41

ciences

-0.40

 amid

-0.40

 saddened

-0.39

 toimi

-0.39

ency

-0.38

POSITIVE LOGITS

 white

2.28

white

2.22

White

1.96

 White

1.92

 WHITE

1.84

WHITE

1.74

 weiße

1.27

白

1.24

 weißen

1.23

 beyaz

1.20

Activations Density 0.186%

references to race, specifically the term "white" and its variations

No Comments

No Known Activations