INDEX

Explanations

phrases that introduce or reference previously mentioned content

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_21/width_16k/average_l0_70

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

rrggbb

-0.73

routeProvider

-0.71

encodeWith

-0.69

VersionUID

-0.69

windowFixed

-0.68

stateProvider

-0.66

TagMode

-0.65

HandlerContext

-0.64

HostException

-0.63

łaszcza

-0.63

POSITIVE LOGITS

 stated

0.62

 faid

0.53

 said

0.52

 already

0.51

said

0.50

 дан

0.50

 regards

0.50

 compared

0.49

how

0.49

 redan

0.48

Activations Density 0.155%

phrases that introduce or reference previously mentioned content

No Comments

No Known Activations