INDEX

Explanations

derogatory terms and statements about individuals and groups

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 kasarigan

-0.66

adaptiveStyles

-0.58

 CURIAM

-0.52

AfterViewInit

-0.52

 bershka

-0.50

 esternos

-0.50

homonymie

-0.47

itinéraire

-0.46

cterms

-0.45

 iſt

-0.45

POSITIVE LOGITS

 Italijanski

0.40

sne

0.36

 shame

0.35

Dam

0.34

Kom

0.34

 Salazar

0.33

render

0.33

Slu

0.32

 blinded

0.32

Activations Density 0.163%

derogatory terms and statements about individuals and groups

No Comments

No Known Activations