INDEX

Explanations

references to specific individuals, particularly in legal or familial contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 15-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_15/width_16k/average_l0_78

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.15.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

HandlerContext

-0.62

="'+

-0.55

 jadx

-0.55

igans

-0.53

 שוליים

-0.50

 المعيارى

-0.49

 sija

-0.48

 stanovnika

-0.48

asty

-0.47

,:);

-0.47

POSITIVE LOGITS

 sibling

0.68

 brother

0.68

 sister

0.67

 brothers

0.60

homonymie

0.57

 CommonModule

0.57

 Infórmanos

0.55

 siblings

0.55

 daughters

0.54

 sisters

0.53

Activations Density 0.132%

references to specific individuals, particularly in legal or familial contexts

No Comments

No Known Activations