INDEX

Explanations

references to specific islands

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-2b-pt-res/layer_1/width_16k/average_l0_102

Prompts (Dashboard)

36,864 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

Personensuche

-0.75

AnchorStyles

-0.65

 تانيه

-0.65

 jScrollPane

-0.63

 مشين

-0.62

 nén

-0.59

InjectAttribute

-0.59

éton

-0.59

 Breen

-0.59

verwijspagina

-0.58

POSITIVE LOGITS

 Island

1.25

Island

1.06

 ISLAND

1.01

 island

0.93

island

0.87

 Central

0.84

Central

0.73

CH

0.70

CH

0.69

endif

0.68

Activations Density 0.073%

references to specific islands

No Comments

No Known Activations