INDEX

Explanations

derogatory and vulgar terms

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

after insults or negative terms

np_acts-logits-general · gemini-2.0-flash

insults and derogatory terms

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

CodeAttribute

-0.72

StoryboardSegue

-0.69

 مشين

-0.68

 cref

-0.68

 referenties

-0.66

riages

-0.65

httphttps

-0.64

FieldBuilder

-0.63

marginVertical

-0.63

WaitGroup

-0.62

POSITIVE LOGITS

bag

0.62

meister

0.49

ofag

0.49

bags

0.48

fest

0.47

 doos

0.46

android

0.46

bucket

0.46

factor

0.45

 hither

0.45

Activations Density 0.326%

derogatory and vulgar terms

after insults or negative terms

insults and derogatory terms

No Comments

No Known Activations