INDEX

Explanations

various types and styles of objects or concepts in a contextual framework

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 27-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_27/width_16k/average_l0_118

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.27.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

 juges

-0.41

 Evangelio

-0.40

 posesión

-0.39

 herido

-0.39

 solidaridad

-0.38

 transparencia

-0.38

 hospitales

-0.38

 Económica

-0.37

 pédagogique

-0.37

 plegable

-0.37

POSITIVE LOGITS

ValueStyle

0.72

Tembelea

0.66

protoimpl

0.65

 autorytatywna

0.64

 uska

0.60

 kasarigan

0.59

NameInMap

0.59

InitStruct

0.58

áne

0.58

ніципалі

0.58

Activations Density 0.566%

various types and styles of objects or concepts in a contextual framework

No Known Activations