INDEX

Explanations

references to controlled extensions in programming or data structures

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-pt-res/layer_2/width_16k/average_l0_67

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.2.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

Jeografia

-0.62

-------

-0.60

BoxShadow

-0.59

tagHelperRunner

-0.59

Spread

-0.57

awtextra

-0.56

 Spread

-0.56

webElementXpaths

-0.55

-0.54

 CallOverrides

-0.52

POSITIVE LOGITS

大

0.83

大

0.54

 controlled

0.54

Controlled

0.53

 Controlled

0.53

대

0.52

@[+][

0.50

대

0.45

controlled

0.44

'\

0.44

Activations Density 0.149%

references to controlled extensions in programming or data structures

No Comments

No Known Activations