INDEX

Explanations

references to "Hell" or related terms in various contexts

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

after "Hell" or "hell"

np_acts-logits-general · gemini-2.0-flash

hell and hel sequences

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

bootstrapcdn

-0.69

 دیکھیے

-0.66

soka

-0.66

ittäin

-0.63

évaluateur

-0.62

 ویکی‌پدی

-0.62

werp

-0.62

 للاسماء

-0.62

cery

-0.60

 Theſe

-0.60

POSITIVE LOGITS

uva

0.60

 Hell

0.56

ishly

0.54

fire

0.53

 demonios

0.53

 peggio

0.53

Hell

0.53

hells

0.52

 HELL

0.49

 devils

0.49

Activations Density 0.083%

references to "Hell" or related terms in various contexts

after "Hell" or "hell"

hell and hel sequences

No Comments

No Known Activations