INDEX

Explanations

various forms and types of classification or categorization

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

some form of

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_131k/average_l0_109

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.31.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

任何

-0.35

multirow

-0.35

any

-0.35

 sekali

-0.35

 vieles

-0.34

 любых

-0.34

mno

-0.33

 makeStyles

-0.33

Any

-0.33

 respectivamente

-0.33

POSITIVE LOGITS

 متعلقه

0.73

しら

0.64

Tembelea

0.63

audiovisuel

0.63

 Normdatei

0.63

 للمعارف

0.62

 بيها

0.57

 خارجية

0.56

httphttps

0.56

 sort

0.55

Activations Density 0.035%

various forms and types of classification or categorization

some form of

No Known Activations