INDEX

Explanations

references to the author or creator of a work

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 37-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_37/width_16k/average_l0_124

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.37.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

 bowiem

-0.56

horabuena

-0.55

 męska

-0.46

تفصیلات

-0.45

 tarko

-0.44

 penghargaan

-0.44

awaran

-0.43

 ocze

-0.42

 rispondere

-0.41

 adelante

-0.41

POSITIVE LOGITS

stander

0.65

standers

0.64

products

0.57

BY

0.56

stolic

0.55

gone

0.55

 virtue

0.54

By

0.54

 means

0.53

BY

0.53

Activations Density 0.108%

references to the author or creator of a work

No Comments

No Known Activations