INDEX

Explanations

references to popular platforms and commonly used items in technology and medical contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 13-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_13/width_65k/average_l0_75

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.13.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

featureID

-0.88

 propOrder

-0.84

OGND

-0.71

octanol

-0.71

webElementXpaths

-0.70

يكب

-0.68

__":

-0.68

PhysRevD

-0.67

 betweenstory

-0.66

"]);

-0.65

POSITIVE LOGITS

 ubiquitous

0.81

ubi

0.71

 universally

0.68

 popular

0.67

 everywhere

0.65

 popularity

0.64

万能

0.63

 ovunque

0.62

 staple

0.62

 universal

0.61

Activations Density 0.200%

references to popular platforms and commonly used items in technology and medical contexts

No Comments

No Known Activations