INDEX

Explanations

URLs and linked resources in the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 15-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_15/width_16k/average_l0_78

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.15.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

 архивлан

-0.66

 ostavi

-0.63

StructEnd

-0.61

sizeCache

-0.61

Personendaten

-0.60

 Meksiku

-0.56

expandindo

-0.55

 oprot

-0.55

DeleteBehavior

-0.54

 Italijani

-0.54

POSITIVE LOGITS

itoneal

0.67

cythe

0.67

$/

0.66

déric

0.63

/}

0.62

{}/

0.61

ADRID

0.61

ftant

0.60

cgi

0.60

enus

0.60

Activations Density 0.068%

URLs and linked resources in the text

No Comments

No Known Activations