INDEX

Explanations

tags indicating the start of sections or specific types of content in structured documents

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

Negative Logits

 arşivlendi

-0.79

 féd

-0.71

 avoient

-0.71

 présidenti

-0.68

 inoxydable

-0.67

 dezelve

-0.66

 pouvoit

-0.66

;;)

-0.66

 légitime

-0.66

arranty

-0.65

POSITIVE LOGITS

 мәкал

0.76

исленность

0.69

WebVitals

0.64

#+#

0.62

enumi

0.61

PhysRev

0.60

+#+

0.58

StandardCharsets

0.58

########.

0.55

Slf

0.53

Activations Density 0.158%

tags indicating the start of sections or specific types of content in structured documents

No Known Activations