INDEX

Explanations

structured data formats and identifiers

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 3-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_3/width_16k/average_l0_90

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.3.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

存于互联网档案馆

-0.71

sidemargin

-0.67

endcsname

-0.65

ChildScrollView

-0.63

reportWebVitals

-0.63

 lenker

-0.62

\{\\

-0.60

IntoConstraints

-0.57

 كومونز

-0.57

帖最后由

-0.56

POSITIVE LOGITS

 génie

0.44

 seorang

0.44

individuals

0.42

 conclusão

0.40

 Individuals

0.39

 istrinya

0.39

Individuals

0.38

 himself

0.37

 drapeau

0.37

 chinoise

0.37

Activations Density 0.587%

structured data formats and identifiers

No Known Activations