INDEX

Explanations

structured code-related terms and directives

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 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

Autoritní

-0.73

 useAppContext

-0.69

XmlAccessorType

-0.68

RenderAtEndOf

-0.67

 propOrder

-0.67

StoryboardSegue

-0.66

contentLoaded

-0.66

 мәкал

-0.64

 ویکی‌پدی

-0.62

hoeddwyd

-0.62

POSITIVE LOGITS

AddTagHelper

0.71

 corresponding

0.50

ViewFeatures

0.46

ValueStyle

0.46

 line

0.46

 hidden

0.45



0.39

 default

0.38

//$

0.37

 NSFW

0.36

Activations Density 0.020%

structured code-related terms and directives

No Comments

No Known Activations