INDEX

Explanations

intensifiers or modifiers that emphasize the degree of a particular quality or state

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

참고

-0.81

 betweenstory

-0.77

 متعلقه

-0.71

GEBURTSDATUM

-0.68

λίου

-0.66

 saites

-0.61

Біографія

-0.60

afficheront

-0.59

 SwiftUI

-0.59

☸

-0.59

POSITIVE LOGITS

 many

0.65

 fewer

0.59

 Sebab

0.57

 more

0.55

 dozens

0.54

 lots

0.52

 everything

0.52

one

0.52

 nearly

0.50

 loads

0.49

Activations Density 0.304%

intensifiers or modifiers that emphasize the degree of a particular quality or state

No Comments

No Known Activations