INDEX

Explanations

repeated or emphasized references to concepts and relationships

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 40-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_40/width_16k/average_l0_125

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.40.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

enterOuterAlt

-0.63

transQ

-0.50

inghouse

-0.48

UserScript

-0.45

AnimationsModule

-0.44

 Pingback

-0.43

Erstellt

-0.42

usual

-0.41

☣

-0.40

ItemBackground

-0.40

POSITIVE LOGITS

 things

1.02

 cosas

0.71

 dingen

0.70

 coisas

0.65

 THINGS

0.63

way

0.59

 people

0.59

things

0.55

 Dinge

0.54

 changes

0.54

Activations Density 0.353%

repeated or emphasized references to concepts and relationships

No Comments

No Known Activations